Abstract
Oftentimes, people have a black and white perspective on the punishment of those responsible for unjust actions. Even though someone’s brain might biologically operate differently than the average brain, they are usually punished equally. In the famous case of Phineas Gage (Macmillan, 2000, pp.47-50) an iron rod entered under his left eye and up through his skull, landing 25 feet behind him. In the process it destroyed his frontal lobe. The post-accident Gage was not the same as the pre-accident Gage based on Dr. Harlow’s observations and the observations of friends and family. We can all agree that the psychological changes that ensued were not directly Gage’s fault.
Biological factors aside, causal determinism can still be applied. Similar to the nurture aspect of the nature-nurture perspective in psychology, a person’s upbringing can play a tremendous role in their future personality or actions. For instance, an abusive or drug-addicted parent is going to somehow have an affect on his/her child’s life. We must take the deterministic perspective into consideration in our legal system and in our own personal judgments of other people. Just as we have got away from stoning adulterers, burning ‘witches’ at the stake, and lynching people because of their color, we must evolve. We must gradually move away from our unambiguous ideas about justice. Causal determinism further clears the lens by which we judge others’ actions and reactions so that we may develop more accurate opinions of each other.
Discussion
There are several types of determinism. The one this essay focuses on is causal (cause-al) determinism—not to be confused with predeterminism, Calvanism, or theological determinism. Determinism is defined as all things in life including actions, thoughts, thought processes, reactions, and behavior have a cause-and-effect determination based on an unbroken series of prior events. The idea is analogically similar to dominoes falling. For instance, and this is just an example for clarity: John Doe has an intense fear of flying. He has this fear because his mother died in a plane crash ten years ago. Therefore, he does not fly out of country and refuses to see his brother who lives in Britain. Perhaps his brother thinks John simply does not want to see him, however due to casual determinism he prefers to stay at home because his mother died on a plane. Determinism is not always a conscious phenomena; it exists in the subconscious as well, or perhaps in both. Jane Doe might be scared of snakes because she had been bitten at age two, but she no longer recalls the actual occurrence. These are all just basic, general examples of determinism.
Specifically, there are different types of determinism aside from what has been discussed previously. Soft determinists often agree with compatibilism, or the idea that free will and determinism can coexist. In this idea, one’s reactions and opinions might have been shaped by past events, but the individual is not chained to his/her past. Essentially it recognizes determinism as a part of free will.
Hard determinists often stand by the argument of incompatibilism, or the belief that the ideas of free will and determinism cannot coexist in one idea or theory. Determinism is absolute and free will is an illusion, and since one’s birth, a series of events were set in motion—a domino flicked or a clock wound, if you will—and this one action has solely affected all thoughts and opinions since. If you perceive you have free will and you act, in reality you are acting because something subconsciously or consciously affected you previously. The problem with hard determinism is that an evil person’s moral behavior can be blamed entirely on the environment or biology of that person. It causes one to disregard moral control entirely because he/she did not feel in control of his/her life and destiny. Some who perform bad actions use hard determinism as a scapegoat. That is not to say determinists are not in favor of punishing criminals.
Indeterminists renounce determinism, believing that past events have no bearing on one’s decision-making or thought processes, and the human psyche consists solely of free will. This idea, like hard determinism is seemingly more extreme than soft determinism.
I will be mainly focusing on the compatibilistic idea of determinism because it is the least radical. It is the golden mean between indeterminism and hard determinism. It is also the easiest to apply to everyday situations and everyday life, and like in most areas of philosophy and psychology, usually the least black-and-white approach yields the best results. In psychology, the overall consensus now—after much debate—is that both nurture and nature play a role in childhood development. This blending of ideas is the approach I will take with free will and determinism.
First, I will discuss the biological perspective of determinism. There are many individual psychological cases relating to biology of the human brain. In the incredible case of Phineas Gage, it goes to show how the physical alteration of the brain can permanently change personality. After the incident, he was “no longer Gage” (Macmillan, 2000, pp.47-50). He did not function like he used to, and for this he never regained his position at work as a foreman. Harlow and others describe him after the incident as unable to plan and being profane. Although some scholars believe the actual description of post-accident Gage was distorted, he was a changed man.
There are other case studies where brain damage has played a role in moral/ethical decisions (Rutigiliano, 2008, p.1). For instance, in the case of Albert Fish, damage to the frontal lobe had disastrous consequences.
An example of a serial killer that had suffered sever injury to his frontal lobe is Albert Fish, better known as the Brooklyn Vampire. At the age of seven he had a severe fall off a cherry tree which caused a head injury from which he would have permanent problems with, such as headaches and dizzy spells. (2) [4] After his fall he began to display many violent tendencies, including an interest in sadomasochistic activities. At the age of twenty he killed his fist victim, a twelve-year-old neighbor by the name of Gracie whom he cannibalized.Albert is just one example of biological determinism of one’s personality or moral actions. More and more studies have concluded that many extremely violent-natured people have something literally wrong with their brain (Rutigiliano, 2008, p.1).
A startling amount of criminals on death row have been clinically diagnosed with brain disorders. A recent study has demonstrated that 20 out of 31 confessed killers are diagnosed as mentally ill. Out of that 20, 64% have frontal lobe abnormalities. (1) [4] A thorough study of the profiles of many serial killers shows that many of them had suffered sever head injuries (to the frontal lobe) when they were children.Other biological predispositions may affect a person’s future judgment, character, or actions. For example it is well known that there is actually a specific gene responsible for the “addiction” complex. Often I hear people say that they are predisposed to becoming an alcoholic because his/her father was. I have also heard the same thing said about the gambling mentality. Granted that it is no excuse for bad behavior, it is still something to be considered.
Now I will discuss the environmental perspective of determinism. The environment of a child at an early age often plays a large role in how they will act for the rest of their life. If their father shows no respect and curses the majority of the time, the child will pick up the same habits. At such a young age, it is hard for children not to follow in their parent’s footsteps, as they are starting with a clean moral slate. They have no immediate perception of what is right and wrong, especially about the abstract, until their parents pave the way. Fear about a reoccurring event in obvious ways affects a person’s lifestyle. If one was too scared to drive a car because of a dangerous wreck caused by a drunk driver that might cause them to stay at home. They would lose social contact with the outside world. The wreck might also cause them to despise anyone who drinks even in moderation. The chain continues. If a person grew up in a low income area with drug-addicted parents, and they literally knew nothing except crime and gangs, having never seen the other side of life, how can we expect them to resort to anything different from what they have seen? Crime, drugs, and violence often is the ‘language’ these people have spoken their entire lives. If no one is willing to teach them a different language—one of love or responsibility—how can we judge them?
Conclusion
Do we, as horribly flawed creatures, have a right to judge others based on something they cannot control? Some of us need to step back and rethink the morality of people’s actions in general. Often it is easy to say, “well if I were in that situation, I would never do that.” In reality we cannot say that in all truth because we would not be ourselves in that situation. We might be someone else, regressing into our primal selves or acting out of character. It is always better to judge oneself before judging others, as we see not the full picture. Causal determinism further clears the lens by which we judge others’ actions and reactions so that we may develop more accurate opinions of each other.
References
Macmillan, M. (2000). Restoring Phineas Gage: A 150th Retrospective. Journal of the History of the Neurosciences, 9(1), 46-66. Retrieved from Academic Search Premier database on Feb. 27, 2010.
Thimble, M. (1990). Psychopathology of Frontal Lobe Syndromes. Seminars in Neurology, 10(3). Retrieved from www.ect.org on Feb. 26, 2010.
Rutigliano, A. (2008). Predestined Serial Killers. Serendip’s Exchange. Retrieved from Serendip’s Exchange on Feb. 26, 2010.
Lois Rogers Medical, E. (n.d). Secret tests on brains of serial killers. Sunday Times, The, Retrieved from Newspaper Source Plus database on March 10, 2010.
Hello, guest philosopher!
ReplyDeleteI find the golden mean argument a bit curious, since you don't really bother to defend indeterminism. I would agree that indeterminism is indefensible, as you indicate with your many examples of clear (if partial) determinism. So it is hard to understand why this is used as a pole position, anchoring one side of the argument, since hard determinism has a great deal more going for it.
For instance, basic physical causality indicates that our actions, insofar as they are caused by brain events embedded in the physical world, descend from past physical states, i.e. are determined. What appears to escape determinism I would account for by learning. We can learn rules and learn from experience, and this is both entirely consistent with hard determinism, (computers can learn as well, at least a little), and provides an explanation for moral controls exercised seemingly "freely".
As you say, hard determinists have no problem with rules and punishments, which serve as a learning/feedback process embodying past lessons in social living learned by our forebears. If a person can't learn and abide by such rules, they may be deemed incorrigible, even psychopathic, and put under communal control (locked up) in liu of self-control. All this can work in a deterministic model provided with learning potential and introspection. And it also, as you write, may contribute to compassion for those afflicted with lack of self-control.
Let me also offer that we still have very faint knowledge of what goes on in development and in the unconscious, and thus what form determinism of whatever kind really takes. You have offered quite a few clues, and the indication (to me, at least) is that the more we learn, the more determinism we will find.
Hi,
ReplyDeleteBravo for jumping in.
Do we, as horribly flawed creatures, have a right to judge others based on something they cannot control?
This is a very important question and I believe we don't. You focus on pathological cases but the argument, I think, goes for most if not all our actions and it may call into question the whole notion of moral responsibility. Whatever our philosophical outlook is it should be clear that our actions are influenced (whether they are determined or not) by an essentially infinite sequence of events over which we may have had very little control.
While reading your text I was reminded of the closing arguments of Clarence Darrow in the famous Leopold & Loeb trial where he very eloquently uses this line of argument. If you have not seen this, you might enjoy having a look (it is a very long text). One quote:
If there is such a thing as justice it could only be administered by one who knew the inmost thoughts of the man to whom they were meting it out. [...] Who could tell how the emotions that sway the human being affected that particular frail piece of clay. It means more than that. It means that you must appraise every influence that moves them, the civilization where they live, and all society which enters into the making of the child or the man!
Hi
ReplyDeleteThis is such an interesting topic because, as you point out, it has immediate practical implications for our approach to justice. Also it forces me into the very intellectual compromises I tend to oppose in religious thought. Free will makes a hypocrite of me.
I too tend towards a sort of compatibilism. There are three reasons for this, all of which cause me some confusion.
JP, there is a great speech at the end of the film Dogville, where the gangster father accuses his daughter of arrogance, for refusing to hold the people who have abused her to the same standards she demands of herself. For just this reason,and speaking as a school teacher, I can't reach the point of believing we can not judge others, because I am simply unable to live well without judging myself, in the sense of having a set of standards I aspire to. There is an important kind of generosity in extending the same potential to others I think, any less demeans us doesn't it?
Secondly, my attachment to evolutionary theory makes me unable to believe in the reachability of absolute truths. The best we can do, I think, is speak of truth in pragmatic terms. So, I believe in science precisely because of the pragmatic use of predictive accuracy, fecundity, elegance etc. It then feels to me as if I must extend the same style of pragmatic assessment to a belief like free will, which is for me the ultimately pragmatic belief. This has always felt like the strongest argument in favour of religious belief to me.
And finally, Burk, if I push myself to the extent of believing in hard determinism, as my reason encourages me to, I find myself in the bind of being asked to believe in the impotency of beliefs, a troubling paradox. By this I mean that if the way I think, believe, assess, speak, and choose is in fact just the working through of a mechanistic process, then I'm not sure what meaning terms like believe can hold within that context, so I'm led into a sort of contradiction.
So, I am thoroughly confused, and led towards positions I'm intuitively wary of. Help.
Bernard
Hi, Bernard-
ReplyDeleteI'm not sure I understand your paradox. We have beliefs, no question about that. They also induce us to do things and take actions that other beliefs wouldn't lead us to. And they arise from information (or disinformation) we have acquired, whether from experience, from instruction, etc. So all parts of the belief operation are compatible with hard determinism.
An example I am particularly interested in is the belief that the US is "out of money". This induces many to think that the Federal government can't or shouldn't spend more to reduce unemployment. People hear this from many sources, think about their personal household debt situations (which are not analogous) and presto, a belief is formed with momentous consequences.
What may be troubling is the prospect of holding beliefs by the relatively mechanical operation of reflecting inputs of media. What role does reason have, or personal idiosyncrasy? Determinism has no problem with your childhood experiences in the Depression having strong effects on your current beliefs and susceptibility to specific media themes. Nor does it have a problem with your ability to sift through all your conscious information, puzzling over discrepancies and focusing on empirical data and sound theories to conclude that the federal government has zero financial constraints, zero solvency issues, and very few tax implications from more spending, in this particular case.
So, for issues that seem to involve free will, I would urge you to break them down as closely as possible, and I think you will find ever more determinism.
Burkhard
ReplyDeleteI'm unsure on the nature of this paradox too, I need to dig deeper, and because I'm hazy on it I'll explain it badly.
It is to do with what I mean when I say I believe something. If all I mean is that my lifeline has produced in me a predispositon to make certain utterances, or mechanical responses, then this is entirely compatable with hard determinism. And indeed, this is where reason leads me.
But, I suspect that I, and pretty much everybody I know, is using the word belief in a different way. So, to take the belief about a governement's finances, let's say I do believe that at a particular moment it would be economically risky for a government to increase its spending. Were that the case, I don't mean just that I have a tendency to utter a sentence to that effect, or to put a particular mark on a voting paper. I mean something more. I mean, amongst other things, that my conception of government refers to some real thing in the world, or that the decision I face when casting the vote will in turn impact on that real world.
In other words I use the word belief assuming a belief is not just a behavioural predisposition, but also has an aboutness to it. Eric will be able to help here, I think philosophers refer to this as intentionality or some such thing.
Now, my very unsophisticated sense of this is that hard determinism would run up against this aboutness. I can retain behavioural predispositions, but not the actual aboutness that reasoning itself appears to require to be meaningful. Hence I seem to be no longer allowed to believe in belief if I believe in hard determinism, which would be sort of paradoxical wouldn't it?
Hope this makes at least vague sense.
Bernard
Hi, Bernard-
ReplyDeleteI see .. you seem to be winding back to the problem of consciousness, which is another matter. Needless to say, I take a "hard" & materialist position there as well, but don't have a wrapped up scientific story on that score, so am just betting. But as you say, reason does lead to hard determinism, which should suffice for this argument.
Bernard,
ReplyDeleteOf course you're right – we're judging people all the time. The question is: in regard to what? I was taking “judging” in its moral sense. And here is something I find very difficult to do: morally condemning people for actions they were maybe helpless to prevent. Not because of determinism (I suspect that correctly analyzed this does not change much) but because of all the unknown elements in their life – I am with Darrow here. On the other hand, I might absolutely condemn the action itself but it's very different from condemning the person.
As for holding people to standards, this is interesting. I have not seen the film you mention and I may misinterpret but the argument seems to be: if you hold others to lesser standards it is because you feel they are lesser persons than you and cannot be expected to do as well – hence the arrogance. Maybe it makes sense in the film but I don't see it this way. Looking at myself, I certainly set personal standards in certain areas, high enough in some cases. But I see that more in terms of personal goals: say, I just want this piece of work I am doing being of this level, and so on. That others may have lesser or higher personal standards – I don't really mind.
Now, in some relationships it becomes a different matter. I have also been a teacher (long ago) and here, of course, you expect your students to achieve certain standards – this being materialized by grading their work, for instance. Or, if I hire a painter to work on my home I expect certain standards to be met. This kind of things.
My issue really is with the kind of facile moral judgments that says: because of his/her actions, this person is morally inferior or deserves punishment (without ever trying to find and understand why). I can't help finding something repulsive in this.
Hi JP
ReplyDeleteYes, I fully agree on the dangers inherent in accusing others of moral weakness when we are unable to understand the circumstances behind them. An attitude of attempting always to first understand is crucial to tolerance.
The standard I have in mind to which we must hold others to if we are to afford them dignity, is the standard that sits at the heart of the free will debate. It is the standard that says, I believe you are capable of making decisions and taking responsibilty for them. Because I base my life about the assumption that this is true for me (the narrative I tell myself about my participation in my own life is critical to its success) then I fele moraly obliged to extend it to others.
At this point, to all extents and purposes, I believe in free will, despite a very strong logical argument that it is incoherent. At this point pragmatism simply trumps reason, which in itself is a highly reasonable stance perhaps?
Bernard
Hi Burk
ReplyDeleteYes, I share your suspicion on the way the consciousness problem will fall. This does not however seem to get to the heart of the problem for me.
It has something to do with the criteria by which reason is itself held to be the supreme measure of the value of a stance. For me, it has something to do with the proven pragmatic value of reason; I don't have any instinct for the sort of mysticism required to elevate reason beyond that. This being the case, pragmatism trumps reason, and hence those assumptions required in order for our faith in reason to work can not in turn be reasoned out of existence. I'm wondering if free will isn't actually one of those foundation values upon which reasoning is perched. Not sure about this, but I'm not willing to let go of this line until I understand it better.
Bernard
Bernard,
ReplyDeleteYou hit on a very interesting point, one which perplexes most theists. Basically, you seem to be saying “my reason and rationality tell me there is no free-will, but my experience, deepest inclination, and intuition tell me I do have free-will-so I must be pragmatic and put my reason aside.”
This raises many questions. First, don’t we have to assume that “reason” and “rationality” only make sense as categories, as states of mind, as ways of proceeding, as ways of evaluating, as ways of deciphering information and experiences, only if we are TRUELY free agents in the traditional sense of those terms and in the way most people understand those terms? If not, then whence “reason” and, frankly, who would care one way or the other what was reasonable?
Second, when you, Burk, and JP talk about “reason” and “rationality” are you aware you are using these terms in completely modern ways when one of the points of the postmodern was to reveal that there is no such thing as an objective “reason” or “rationality” and that “reason” is always and only a tool in the employ of our larger narratives or world-views? In other words, what we think “reasonable” and “rational” are imagined concepts and as such, they rely upon what comes before them which is our faith, our passions, our deepest held beliefs, our world-view as it were. It is not as if there is something out there, some universal force, like gravity, called “reason” that we simply latch on to and use. “Reason” as you are using it has a historical and philosophical genealogy, which I think you guys would be hard-pressed to defend. It is simply assumed and never defended, which is partly why there is so much talking past one another in these conversation.
Third, what does it say about a view of “reason” where one has to basically say “what my reason leads me to believe in this area of free-will is so false to my experience and deepest intuition about my own sense of who I am in the sense of being free, that I must not believe it to be true.”? Does such a view even seem reasonable or rational, regardless the pragmatism?
Conversely, if the Judeo-Christian God exists, then free-will makes sense and is both reasonable and rational. It also has the added benefit of actually squaring with our experience and deepest sense of ourselves. No small thing, that.
Darrell,
ReplyDeleteWith respect to the following: "Conversely, if the Judeo-Christian God exists, then free-will makes sense and is both reasonable and rational."
I'd be interested to see your case for that. I say this because I have been wrestling with the concept of free will WITHIN theistic traditions--as part of my ongoing work on the doctrine of hell and the question of the coherence of the supposed "free choice" of damned to exist in a state of hell.
Hi, Darrell, Bernard, JP-
ReplyDeleteI think the "reason" why reason is so hard to reconcile with our immediate experience on the subject of free will is not hard to find- pure ignorance. Specifically, we are profoundly ignorant of the origin of our own promptings, motivations, decisions. They "pop" into our heads and seem free, not to say god-given. This is as far as our immediate experience goes, and it is understandably shrouded in a nimbus of wonder. But our capacity for introspection is notoriously lousy, so one person's wonder is another's ignorance of causes that are determinisitic- on the theoretical plane as well as on the plane of detailed analysis as far as that has actually been done.
It is ignorance that we are well to be glad of, but still, no reason to glamorize it.
Eric,
ReplyDeleteI have not honestly thought through how free-will impacts or affects the doctrine of hell. It seems to me it would all depend upon a host of other factors such as one’s views regarding predestination, hell in general, salvation, the nature of the Trinity, and eschatology.
My case for free-will making sense if God exists in the context of the post and this current conversation is pretty simple. I think the orthodox view of the Christian God is that such a being is truly personal, free, and an acting agent. If we are made in that image, then we have that aspect as well—although given the fall the image is broken. I thought that was pretty conventional orthodox teaching.
What am I missing?
Hi Bernard,
ReplyDeleteIs this really the case that free will and determinism are mutually exclusive? I find it difficult to make up my mind on this but one thing seems clear enough: we make choices all the time, real choices. We ponder the facts, we consider different arguments, we try to anticipate consequences, we draw upon our experiences, and so on. And, the end, we make some choice or another. This is real, not some kind of illusion – there is a level at which this process goes on.
Now, in what respect are these choices not-free? They would be such if there were some constraint on them. And, of course, there are: we cannot consider facts that we don't know about, we are influenced by our past experiences, and so on. But these would be there under any free will definition anyway, so they don't really matter.
In other terms, what is missing for these choices to be free? What could we add to this process to satisfy the definition of “free”? I don't know. I don't see what it could be.
We oppose free will and determinism by saying in effect “our choices cannot be free because our molecules are not”. But isn't it the same as saying “we cannot be alive because our molecules are not”? Or “we cannot digest food because our molecules do not”? Freedom may arise at a higher level of organization, just like life and digestion.
I confess that I don't see very clearly here. What do you make of this?
Determinism is the idea that, unless some external forces affect it, the state of a system determines all its future states. At the absence of external forces a system can only evolve in one particular way, according to determinism. On naturalism the universe, or all physical reality, is not subject to any external forces, and therefore all that happens is predetermined by its previous state. Nothing could be different than how it in fact is. The whole history of the universe in all its details was codified at the Big Bang. “It was written” as primitive people use to say.
ReplyDeleteDeterminism contradicts free will, because “free will”, as normally understood, entails the reality of making a choice, i.e. in acting in one way instead of another that was also possible. But on naturalism and determinism there are no other possible ways; one could not have acted differently than how one in fact did. Therefore there are no choices, and thus there is no free will, nor personal responsibility.
In my judgment determinism is a remarkably bad idea. Since the demise of classical physics about a century ago there is zero evidence for it. (If the reader can suggest even the slightest argument for determinism I’d like to hear it.) At the same time determinism implies conceptual problems galore, and not only in relation to free will. Remarkably it even contradicts science, for according to quantum mechanics the current state of physical system only determines the probability distribution of its future state, but not its future state. Given all that it strikes me as amazing that so many people still cling to determinism. I think it shows how little people really care about evidence when it does not suit their preconceptions.
Hi, Dianelos-
ReplyDeleteDeterminism is fully compatible with personal responsibility, as I have previously explained. Responsibility means that one abides by a rule or general aim. That requires that one understands the rule, and adapts one's behavior to it, and in general that one can learn from instruction or experience. All this is possible with deterministic systems.
Determinism is fully compatible with making choices. It only says that those choices have an array of influences that have made them what they are, (including random phenomena), not that they drop out of some magical ether, or arise from a non-causal, non-physical "self".
Determinism is fully compatible with quantum mechanics and other disiderata of modern physics. Determinism (in a modern sense) isn't knowing what the future holds in the form of some Laplacean machine, but simply adhering to physical causality, such that the output of one's mental and other processes is fully determined by the prior set of physical events, deterministic or not. It is in essence just a way of denying supernaturalism and soul-ism as accounts for human agency. I hope that clarifies a bit. My definition may be new to you.
Hi guys
ReplyDeleteDarrell. Yes, I agree with much of what you say. Reason is itself a system of our own creation I think, and therefore we need to assess reason according to some criteria, and whether these criteria are themselves reasonable becomes a difficult question, there's certainly room for destructive circularity there. An important question though, and I'd be reluctant to leave it just at the 'well, reason is subjective too' level, because clearly finding the types of reasoning that work best, at least for oneself, is important.
Perhaps we can, for social purposes, promote those kinds of reasoning which we share with those around us. The extent to which some forms of reasoning are universal then becomes interesting. Theism doesn't help me with the free will problem much I'm afraid, but perhaps I'm missing something here.
JP, for me determinism and free will do run up against each other. Free will requires that I face a choice and can actively pick a path, that in essence both paths are open to me. Not just that it feels as if they are, but that both are actually available. That means in principle that an onlooker would be unable to anticipate my choice. But it also requires that this inscrutability is not gifted to me via a random component, as this is free but hardly willful. Now, determinism as defined by Burk, (which is the right definition I think) says that the outcome is the result of the physical inputs, allowing that some of these inputs may be genuinely random.
So, it seems to me, one can have will, behaviour guided by the weighting of reasons and influences, or freedom, behaviour with a random component, but one can't have both. This appears to be a problem whether one is a theist or an atheist, free will just can't exist, the concept is incoherent.
And yet, in order to live, I must simply ignore this conclusion I think, which may be Burk's point about ignorance. I am interested in whether this has something important to tell us about reason itself, as Darrell is getting at. I'm uncertain.
Bernard
Hi Burk,
ReplyDeleteYou write: “ Determinism is fully compatible with personal responsibility, as I have previously explained. Responsibility means that one abides by a rule or general aim. That requires that one understands the rule, and adapts one's behavior to it, and in general that one can learn from instruction or experience. All this is possible with deterministic systems.”
If from the very beginning of the universe it was determined that on the planet Earth on October 18, 1939, a baby would be born to be called Lee Harvey Oswald who on November 22, 1963, would kill President Kennedy – I really don’t see in what sense the fact that Lee Harvey Oswald understood the rules makes him personally responsible for his actions. He couldn’t help it knowing the rules, and he couldn’t help it not abiding by them, and so he couldn’t help it shooting the President, as much as a mechanical clock can’t help it moving clockwise. All he did was the result of impersonal forces moving things around in a blind mechanical manner. Thus he is not more responsible for the death of the President, than is a boulder that rolls down a mountain and falls on a person killing her.
“Determinism (in a modern sense) isn't knowing what the future holds in the form of some Laplacean machine, but simply adhering to physical causality, such that the output of one's mental and other processes is fully determined by the prior set of physical events, deterministic or not.”
Perhaps you mean that there are indeterministic elementary events taking place in our brain, but that our brain’s output is causally fully determined by such elementary events. If that’s what you mean then 1) you are painting a non-deterministic picture of the world after all, and 2) you are still contradicting science according to which *all* physical interactions and thus all causal relationships are of an indeterministic nature.
And, in any case, if you don’t like the old definition of “determinism”, rather than redefining it you should coin a new word that means what you wish to convey. Determinism means that the past fully determines the future, and that’s that. And, therefore, that given that there is one past that there can only be one future. To avoid confusion anything else shouldn’t be called determinism, but something else.
“ It is in essence just a way of denying supernaturalism and soul-ism as accounts for human agency.”
Please observe that naturalism does not imply determinism. One can very well believe in naturalism and in the causal closure of the physical and in the absence of souls and of supernatural effects – without being a determinist. One can, as quantum mechanics says, believe that the state of a system determines the probability distribution of its future state but not its future state. Or, in other words, that a system may naturalistically (i.e. in a fully mechanical manner and without being guided by any purposeful will) evolve in one of many different future states. (Unfortunately that view is also incompatible with free will and personal responsibility. That's why free will is evidence not only against determinism but also against naturalism.)
Hi Dianelos
ReplyDeleteAre you able to clarify something for me, please? You say that in science all physical interactions are of an indeterministic nature. Although my understanding of this is inexpert, I thought that at the macro level decoherence washes out quantum effects. So for example if a bus were to collide with me at speed, there is something essentially deterministic about the macro fact of the transfer of energy at collision. It's going to hurt, deterministically so, irrespective of variability in the ways that damage may occur.
If this is true, then in principle the same aggregation effects might hold in the mind. The output, the binary decision to take the risk and cross the road or not, might have the same deterministic quality as the resulting collision, being an inevitable result of the inputs that hold at that point in time.
I'm also keen for someone to rescue me from the free will bind, and you seem to happily believe in free will. How can a decision be free without being random (and hence not willful)? I'm still struggling to even imagine what that might mean.
Bernard
Hi Bernard,
ReplyDeleteYou write: “You say that in science all physical interactions are of an indeterministic nature. Although my understanding of this is inexpert, I thought that at the macro level decoherence washes out quantum effects.”
Quantum coherence is a special state which is very difficult to maintain, but which has little to do with the discussion at hand. According to QM all systems evolve probabilistically (i.e. non-deterministically), but for larger systems the probability distribution tends to be concentrated close to one state which approximates the “deterministic” result one would expect on classical mechanics. It’s a little like tossing a million coins; one will quasi-deterministically get very close to 50% of heads.
“So for example if a bus were to collide with me at speed, there is something essentially deterministic about the macro fact of the transfer of energy at collision. It's going to hurt, deterministically so, irrespective of variability in the ways that damage may occur.”
Actually that’s a good example. If a bus were to collide with you at high speed there is a extremely small but non-zero probability that you would survive without noticing a thing – because of quantum indeterminancy. In this context see also “quantum suicide”.
“I'm also keen for someone to rescue me from the free will bind, and you seem to happily believe in free will. How can a decision be free without being random (and hence not willful)? I'm still struggling to even imagine what that might mean.”
I, on the contrary, fail to understand why so many people appear to be perplexed with free will. Free will is exactly what common people mean by that concept, namely that we often have to choose one among a number of alternative causes of action, and we pick one by our own free will. We could have picked something else, but we end up picking one particular action, sometimes after quite some internal deliberation and of being torn between various forces such as our desire to do the right thing, our desire for pleasure, our fears, our envy, our shame, our desire to impress others, etc. The exercise of free will is a very common every day experience for all of us. I really fail to understand what the big problem with it is.
For example you are asking “how can a decision be free without being random?” but I don’t understand what the problem is for the answer strikes me as being trivial: A free decision is not random because it was chosen by me, and I am a feeling/reasoning/purposeful person and not some kind of random source.
Could you please explain to me what problem you see with free will, for sometimes I fear I may be missing something important here.
Hi Bernard,
ReplyDeleteI am not trying to “save” free will or anything but it really seems to me that the dichotomy between free will and determinism does not add up.
You write Free will requires that I face a choice and can actively pick a path, that in essence both paths are open to me. Not just that it feels as if they are, but that both are actually available.
But, in a real sense, we can choose a path and both are available. There is no mystical “deterministic” force swaying us one way or another – this is looking at the situation backward. It may seem after the fact that the path non-chosen was not “really” available but it is because we choose the other one. The fact that only one path will be chosen does not make the choice forced.
When I think about determinism I have this image in my mind of some “deterministic force” enforcing its will on my decisions. But the image is wrong - there is no such force. However, it's quite difficult get rid of it, but we have to.
Predictability, I think, is not a problem either. Imagine for a moment that there is Free Will. An observer outside space-time would be able to look into the future and see what any Free decision will be.
Hi JP
ReplyDeleteI'm not seeing this yet. Let's say I'm walking into a supermarket and there's someone collecting for a charity outside. Do I give them the ten dollars I'm carrying?
I agree that giving and not giving are both apparently possible. No great outside force reaches into my pocket. Nevertheless, I do decide, one way or the other. What then underpins my choice? I can only see two options. One is the running of a sort of algorithm, taking into account all my beliefs, desires, and inclinations, essentially a physical output determined wholly by physical inputs, in principle predictable from within space time. This is determined in the sense that a computer making choices in a chess game is determined. It is not a free choice in this sense, our decisions are the playing out of a programme.
Or there is some random component at play, so that somewhere within the weighing up of options there is a genuine uncaused leap in there, the way we think of the collapse of a quantum probability wave function perhaps. This is free in the sense that it can not be accurately predicted within space time, but it is however not willful.
So I'm stuck with the thought that what is willful can not, in a important sense, be free. Unless there is a third option that I'm missing.
Bernard
Hi, Dianelos-
ReplyDeleteI agree that Laplacean determinism is a non-starter- the empirical evidence refutes it, and I don't think anyone defends it. But I don't think that is what the guest blogger had in mind with his post, either:
"If you perceive you have free will and you act, in reality you are acting because something subconsciously or consciously affected you previously."
This is his (her?) presentation of hard determinism, and nothing else tells me that the model demands Laplacean determinism as well. If you could suggest a better term for the naturalistic adherence to causality that is not Laplacean determinism, that might be helpful. I don't think it is "soft" deteminism or compatibilism, since here the guest states "the individual is not chained to his/her past".
So I am curious what you mean by free will, after all. "Free will is exactly what common people mean by that concept, namely that we often have to choose one among a number of alternative causes of action, and we pick one by our own free will." Who is that "we"? Who is that "one" homunculus who lives outside causality and can "choose", based on what, for what? I don't think your model makes sense if you interrogate it closely, as Bernard alludes to.
By naturalism, the self is a coalition of processes arising from the brain which integrate past experience, developmental and genetic givens, and yes, quantum chaos, experiencing an ongoing lived experience where it makes choices all based on real influences/causes, whether rational and reasoned, or deeply unconscious. Our recent election in the US was a prime example of all these influences at work.
"I really fail to understand what the big problem with it is." The problem is that what you write fails to really ask the key question of what is doing the choosing. You portray choosing as a quite deterministic process, full of reasons and influences. What you state in your explicit examples actually agrees with a deterministic/causal model.
"A free decision is not random because it was chosen by me ..." Yet it might be random, by virtue of flipping a coin, or being on the knife's edge of some quantum event in the brain. While watching the recent baseball games, I was struck by the randomness of the hitter's success. The contest is so finely balanced that maybe a few nerve cells, perhaps a stray firing here or there is all the difference between a successful hit and a ground-out. Yet we cheer them on for their unearned talent, as well as for their hard work and training to attain the high (though still random) odds they do.
At any rate, the real question is- what/who is that which is doing the choosing, and how is that process insulated, in your model, from causal reality/closure? And if so, what else does it possibly have to go on to choose whatever it "wants" to choose? Causal closure and naturalism have to be, at this point, the default hypothesis of how living organisms operate, so the burden in on the soul-ist to come up with a defense of (and evidence for) such insulation that goes beyond the lack of knowledge we routinely have of our inner workings, however magical that seems.
Hi Bermard,
ReplyDeleteIt is as you say: our decision process is algorithmic (in some general sense). But what does this have to do with whether it's free or not? I know we're used to think about this issue in a particular way and assume that, somehow, algorithmic is antithetical to free. But I would suggest that, perhaps, a change of perspective is called for. What, precisely, is this elusive quality that our algorithmic decision process is lacking that we cannot call it free? The more I try to identify it the less it seems to make sense. Isn't this somewhat reminiscent of the élan vital of old? Some élan libre perhaps?
I think we should also try to see determinism (of whatever variety) from another perspective. When thinking about determinism, isn't there this idea that our actions are “decided in advance”? And, of course, seen from this angle freedom goes out the window: that would mean that, although we want to do A we are forced to do B because it's preordained. But this is not how things stand: there is no prior decision about anything.
So, we tend to conflate “determined” with “pre-ordained” and “predictable” with “forced” but this is, I think, mistaken. If we get rid of these ideas, perhaps we can also realize that nothing important is missing.
JP
ReplyDeleteI agree with this. Is there something important missing, becomes the big question.
Whether an algorithmic process can be free is, to some extent, a matter of linguistic preference. It depends upon what one means by free.
What seems clear to me is that the 'I' that does the choosing must, as Burk suggests, be removed from the equation. So, it seems we are free in the same sense that a chess playing computer is free. My question is whether such freedom is missing something important, an ingredient that is key to my narrative of self.
And on this point I am unsure, although I tend toward a yes response. It is exactly to do with this conflating you speak of. Were I to face a choice, and feel I had moral freedom in doing so, but next to me stood a person who was able to unerringly commit to paper the outcome of my decision before I made it, would I feel reduced by this? How hard would my narrative be shaken? (I use this scenario in an upcoming book so it's become fascinating to me).
Part of me answers, as you do, that all is required is a shift of perspective, and less is lost than we imagine. And yet when I try to unpick the basis for values I hold, how I aim to treat others, how I respond to their decisions, or even what I mean when I say I believe something, I do feel as if things start to unravel under this scenario. I need to consider this more I think.
Bernard
Hi Burk,
ReplyDeleteYou write: “I agree that Laplacean determinism is a non-starter- the empirical evidence refutes it, and I don't think anyone defends it.”
Well, surprisingly enough according to a recent survey about 60% of philosophers believe in compatibilism and hence in determinism (see: http://philpapers.org/surveys/results.pl ). Given how bad an idea determinism is, my interpretation is that unfortunately even the beliefs of philosophers have more to do with fashion, or with the image philosophers desire to project, than with evidence. Perhaps Eric can suggest a more charitable interpretation.
Compatibilism actually redefines the concept of free will as expressing that one is not restrained by external forces from executing one’s will. But one’s will itself is still fully restrained and determined by the past. As Schopenhauer concisely put it “Man can do what he wills but he cannot will what he wills”. In my judgment “compatibilist free will” is a shameful philosophical red herring, for the concept of “free will” clearly refers to a property of will and not a property of action. A person entombed up to his chin in cement has very little *liberty of action* but has as much *free will* as anybody else.
“If you could suggest a better term for the naturalistic adherence to causality that is not Laplacean determinism, that might be helpful.”
It seems to me that the naturalist should take science at face value and accept that physical systems can evolve into many different future states. Thus the naturalist should reject determinism in favor of indeterminism, i.e. that physical causality is probabilistic and not deterministic.
[continued in the next post]
[continued from the last post]
ReplyDeleteTo my definition that “Free will is exactly what common people mean by that concept, namely that we often have to choose one among a number of alternative causes of action, and we pick one by our own free will.” you respond:
“Who is that "we"? Who is that "one" homunculus who lives outside causality and can "choose", based on what, for what? I don't think your model makes sense if you interrogate it closely, as Bernard alludes to.”
First of all, let us note that there are two issues on the table here: One, what the concept of “free will” means, and, two, how free will fits within one or the other ontology. My point above was that the concept of free will, which describes a major dimension of our experience of life, is completely clear.
As to your questions above, on the ontology of theism the answers I’d give are as follows: We are transcendental spiritual beings of a personal nature, who in our current condition experience life in a physical environment to which we are closely associated and which severely limits us. In the context of free will, our physical environment (which includes our brain) determines the probabilities of how we shall choose, but does not determine how we shall choose, i.e. within that probabilistic restriction we still enjoy a finite measure of absolute freedom vis-a-vis one particular choice. Vis-a-vis our whole life we enjoy a very great measure of freedom, because each individual choice affects our physical environment and thus how it will restrict our future choices. Thus we are a little like the captain of a sailboat: at each instance the conditions outside and inside the boat severely limit one’s steering, but on the whole one is completely free to steer the boat to one harbor or to another.
Conversely, what would the answers on naturalism be? Rejecting determinism at least allows for the reality of choices. But, on naturalism, “we” can’t be anything else than our brain, or what our brain does, or what supervenes on what our brain does. The problem here is that even though our brain can evolve into different future states of making a choice (and thus choices are real), according to quantum mechanics which of these future states will obtain is a matter of pure chance (there is nothing in physical reality that will cause one of the other future state to obtain). Take again the example of Lee Oswald, and assume that his brain was such that the probability of it evolving into the “I will now shoot to kill the President” state was only 1%. That his brain did evolve into that state was a matter of chance, and thus Lee Oswald (or his brain) cannot be held accountable for that bad piece of luck, nor can it be coherently said that Lee Oswald (or his brain) “freely chose” to shoot to kill the President. Rather it was pure chance that made the “choosing”, which is the same as saying that nobody and nothing actually chose for Lee Oswald to shoot to kill the President. So, on naturalism, even when rejecting determinism, free will remains an extremely problematic concept.
[continued in the next post]
[continued from the last post]
ReplyDelete(Incidentally, I fully accept your description of what on naturalism the “self” is, namely that the “self is a coalition of processes arising from the brain which integrate past experience, developmental and genetic givens, and yes, quantum chaos, experiencing an ongoing lived experience where it makes choices all based on real influences/causes, whether rational and reasoned, or deeply unconscious”. All of what you describe may explain why the probability distribution of the future state of the brain is as it is; but it does not solve the problem that after all has been accounted for the actual decision is still one of pure chance beyond the brain’s control, and indeed beyond the control of the entire physical universe.)
“At any rate, the real question is- what/who is that which is doing the choosing, and how is that process insulated, in your model, from causal reality/closure?”
The only argument I know of against the theistic account of free will is the so-called “interaction problem”. That argument was a good one when people, as per classical science, thought that physical reality is deterministic, and thus that any effect from a transcendental realm would become apparent as an anomaly. As I have explained previously in this blog (see http://thepietythatliesbetween.blogspot.com/2010/06/what-is-naturalism-part-v-alternative.html?showComment=1278660617361#c883817128621853447 ), quantum physics removes that problem for it is now possible not only for humans to exercise their free will without violating the causal closure of the physical, but also for God to exercise His/Her special providence (in creating life on Earth, affecting human history, etc.) without violating the causal closure of the physical either. (As an aside, I find it remarkable how modern science is systematically removing conceptual problems from theism and piling conceptual problems on naturalism.)
Dianelos
ReplyDeleteThe free will you describe is an entirely random process. So free yes, but hardly willful. I'm not sure why you insist upon ignoring the central problem here.
Bernard
Hi, Bernard
ReplyDeleteI think that Dianelos has actually addressed the problem of the self.. "We are transcendental spiritual beings of a personal nature, who in our current condition experience life in a physical environment to which we are closely associated and which severely limits us."
Sounds more like Dianetics than rational thought, but there it is. The "me" is not co-extensive with the body, but something else. Which raises endless problems of what it is, how it interacts with the physical substrate, and why it cares about whatever it cares about. Soul-ists devote enormous literatures of abstruse terminology to the "interaction problem", with no progress or answer, while actual scientists are slowly closing the circuits that make up our mental processes and, in my estimation, will be able to exclude dualism in a matter of decades.
Burk
ReplyDeleteI think the bit that confuses me is how this severely limited self makes the decisions they do. To give to charity or not to give? Once the acknowledged constraints (expected outcomes, values, income etc) are dealt with, what is there left for this self to do?
All I can imagine is that it can give some random twitch that knocks the probability function down one path or another. Anything else, any reasoned or motivated aspect to the decision, is already accommodated by the constraints, surely. If the limited self is in any sense reasonable, then it is constrained by all those things that contribute to a reasoned decision. And so Dianelos' sailer at the tiller becomes unmotivated and spasmodic. I don't think he has addressed this yet, and I'm interested in how he sidesteps the problem.
Bernard
Hi Bernard,
ReplyDeleteYou write: “The free you describe is an entirely random process.”
Not at all; I wonder where you got that from. Free will *as experienced* is certainly not a random process, and that’s why we hold ourselves and others responsible for their actions. Our experience of free will makes perfect sense on theism, according to which we are transcendental beings strongly associated but not fully subject to the physical dimension of reality, and possessing sufficient freedom to ultimately guide our lives this way or the other. And this theistic picture is entirely compatible with scientific knowledge.
It’s only on naturalism that one must assume randomness. Here, as in theism, when faced with a choice, forces such as beliefs, desires, and fears, tug and push one way or the other thus causing a probability distribution among the alternative choices. But whereas theism posits a sovereign/creative/transcendental “I” that makes the final call, on naturalism there isn’t such an “I”, and the final call *must* be something random (while respecting the probability distribution). But if the final call is caused by some mental “random twitch” then, on naturalism, one is clearly not personally responsible. And as, on naturalism, all other participating forces (beliefs, desires, fears, etc) are either caused by external factors or else ultimately the result of one's past random choices, one is not personally responsible for these forces either. On naturalism then one has zero personal responsibility for one’s choices. Thus naturalism makes a joke of one of the foundations of human society and indeed of human relationships, which is the thesis that people are responsible for their actions. On naturalism there is no transcendental “I”, and thus we are just puppets being moved around by a sea of mechanical and blind physical forces. And if it feels otherwise (and it does feel otherwise) then it’s just an “illusion”.
Conversely, on theism, one is greatly responsible for one’s actions, for one is also mostly responsible for the other participating forces, such as beliefs, desires, fears, etc, which are themselves to a large degree the result of one’s past personal choices.
“I'm not sure why you insist upon ignoring the central problem here.”
I am not ignoring it, I am not seeing any problem in the theistic account of free will I gave. Observe that there is nothing “random” in it. Do you see any other problem with that account? Either that in some sense it does not comport with our subjective experience of free will, or that in some way it contradicts scientific knowledge?
Dianelos
ReplyDeleteHere is what I think you are ignoring:
You say the transcendent 'I' makes the final decision. Let's say I accept such a thing exists, as per theism, and this is exactly what it does. Okay, well on what does the transcendent self make its call? It faces exactly the same problem the naturalistic self does, either its final call is algorithmic, pulled by the various factors under consideration, or it is a random collapse of the probability function, as you describe it.
And you are right, this does make a mockery of free will.
This is not a small problem, and as far as I can see is widely accepted by philosophers. Do you have a solution to it? I am not trying to win an argument here, I would love it if you could defeat this conundrum.
Bernard
Hi, Bernard-
ReplyDeleteThat is a very interesting point. I guess that a transcendentalist would say that the transcendent soul is (somehow) aware of all the earthly reasons you indicate, implicit in its connection with the body, but in addition has access to other reasons, such as contact with god, an angel or two, and all the mystical emotions that are the deepest and most profound motivations for human endeavor, and not at all of physical origin whatsoever, and which atheists can't even comprehend, etc. and so on.
So it would be contact with the divine that is specially part of the transcedent soul hypothesis which supposedly reaches beyond the earthly reasons accounted for in your philosophy, so to speak. This could take the form of morality "written in our hearts" as well, insofar as that is not baked in the DNA cake, but exists mystically.
Burk
ReplyDeleteSure, but whether the determining factors are earthly or transcendent doesn't appear to add any freedom into the mix. Freedom seems to require genuine randomness, which is why Daniel Dennett calls this a freedom not worth having.
Bernard
Hi Bernard,
ReplyDeleteYou raise the issue of a predictor that succeeds in writing down in advance the outcome of your decisions and I agree this is a fascinating idea. This is similar to some variants of Newcomb's paradox, one of the most confusing of the kind, I think. The key to these scenarios may be trying to figure out how the predictor can make his predictions.
First, the predictor may be an outside observer looking at our universe (or reality) from some other place without otherwise interacting with it. (This does not fit your scenario because in this one you're not aware of the predictions.) This is a simple case: the predictor may run a fast simulation of our universe or, equivalently, wait until events unfold or, again equivalently, step outside of space-time and observe it as a complete and finished object and simply look. No problem here.
Second, the predictor may try to run a simulation (or do the necessary computations “on paper”, which is the same thing) within our own universe. As he is now part of the universe, his simulation must include himself running the simulation, and so on, recursively. This amounts to executing an infinite number of discrete actions in a finite time. I think we can prove that in a reality in which this is possible, causality would completely break down making predictions impossible (using for example the Thompson's lamp paradox).
We may argue, however, that in order to predict your move it is not necessary for the predictor to simulate himself. Alternatively, then, consider what happens if he tries to predict his future prediction. In this case, I think, we get the infinite recursion and the impossibility of predicting.
Third, the predictor might somehow be able to step out of space-time, observe it from the outside and see what your decision will be. There are logical problems here and I am not sure this is coherent. For example, to consider the totality of space-time we must let it run its course: therefore it does not yet exist for the predictor to observe (we also have to use different times and it becomes very confusing). But never mind that and let's see where this leads us. Part of what the observer sees is its own future prediction. This prediction could not have been determined by an observation he is only now making, doesn't it? It means it could be wrong. In this case the predictor goes back in space-time and makes a prediction different from the one he observed him making.
Now, two things. By making a prediction based on his knowledge of the future, the predictor modifies the very space-time he based his decision on, thereby invalidating the process. Second, by having the possibility of making a prediction different from the one he sees in the future, the predictor satisfies the basic requirement for free will: he proves that he could have done otherwise – by doing so. The scenario, instead of refuting free will, actually appears to require it.
jp
Bernard, JP-
ReplyDeleteSo it seems like you are saying that since god is omniscient and knew all along what Adam and Eve would be doing, the putative grant of free will was/is meaningless- their sin was foreknown, not to say foreordained, precisely because it wasn't free at all. Makes sense to me.
Hi JP
ReplyDeleteI agree that putting a predictor within a universe capable of predicting that entire universe creates the logical problems you outline.
That said, if all one is predicting is the behaviour of another, putatively free actor, the problem may not occur. I can predict that if a ball is released at the top of a ramp it will roll down it. The question becomes, in essence: are we, like the ball, simply responding to physical stimuli, or is there something more to our progress through life than this?
The ball might hit an obstacle on the ramp and face the choice of bouncing left or right. Is this choice, where both paths are open to it, and potentially extremely difficult to predict, different in kind to the choices we face?
One way our choices are different is that we are aware of these choices, and construct narratives about them, which in turn become part of the conditions under which the choices are made. So there is a self referential loop of sorts going on. Furthermore the narratives we construct about others contribute to the way they bounce, and vice versa. This is perhaps the minimum requirement for a meaningful system of morality, and is consistent with a deterministic reality.
The free will theists appear to crave extends beyond this type of self referencing. Specifically they claim a will that is neither physically determined nor random. And this, I contend, is logically problematic.
Bernard
Hi Bernard,
ReplyDeleteI don't know how much we can eliminate logical problems by reducing the scope of predictions. We're talking something immensely complex here, a task requiring to know the target in the most minute details. Obtaining this information seems to require the degree of interactions between the predictor and the target that would force the predictor to take himself into account, thus creating the usual self-referencing problems. In any case, I will have to think more about this.
Some time ago I was reading something by Dennett on free will. At the time, I was somewhat under the impression that he was, by logical trickery, trying to wiggle his way around the “obvious” impossibility of free will. Or that he was arguing for a poor man's version of free will, the best that we could have but “obviously” quite short of the real thing.
But now that, thanks to this blog, I have had the opportunity to revisit the issue, I tend to think that he was absolutely right. Not that I minded – I was perfectly comfortable with the idea of free will as a complete illusion. But now, on the contrary, it rather seems to me that it is the alleged incompatibility between predictability and free will that is an illusion and that the naturalistic account of free will (that I outlined in a previous comment) IS the real thing, and a perfectly good one at that – nothing missing whatsoever.
You are perfectly right, I think, in your estimate of the problems with the theistic idea of free will. It does not seem to add anything beyond the naturalistic account, except some fuzzy, very vague, decision making capability that is neither algorithmic nor random. So, what is it?
jp
What I have argued above is that determinism (in whatever form) is not an obstacle to free will – that the two are essentially independent.
ReplyDeleteHowever, this does not imply that we are actually making rational decisions or that we have moral responsibility. It makes theses possible but not necessary. It may very well be that what we take as rational decisions are often simply rationalizations of unconscious desires or impulses. If this is the case moral responsibility remains a very fuzzy concept.
Hi Bernard,
ReplyDeleteThanks for insisting on this issue. I now think I may (finally) understand what the problem you see is, and also why I didn’t see it before. What follows is a long post, but I hope I will be able to explain how I see things, and also give you a lot of opportunities to find something wrong with it.
Before commenting on the issue of free will, I’d like to clarify a few things:
First of all, when I talk about “theism” I am talking about my own understanding of theism. Theists, as is evident, don’t always agree among themselves. I am here defending the in my judgment strongest version of theism, which I judge to be quite “classic” and “orthodox” – but which qualification other theists may disagree with.
Secondly, when I talk about the “sovereign/creative/transcendental I” I don’t mean the whole personal self, but just what makes the personal self free. It is one element of the self, which I shall henceforth call the “freely choosing element”. For me being a person entails much more, and indeed entails many of the “forces” related to free will in that they shape the probabilities of how one will choose. We have already discussed a few types of such forces, such as beliefs, values/desires, and feelings. While thinking about this issue I saw that there are other types of forces, including, significantly, one’s character. It is clear that all other factors being the same a person who is courageous will probably choose differently than one who isn’t. Another important factor of course are the external circumstances; the state of affairs one finds oneself in and in which one must choose a course of action; one cannot neatly separate oneself from one’s experience of the external environment. The original problem though remains, for all such forces only shape probabilities, but do not make the final call. On the other hand all such forces belong to the personal self too, they are part of ourselves, and they are all of a transcendental nature. The fact that for many of these forces there exists also a physical analogue does not make them less transcendental; this fact only implies that these forces are closely associated with the physical dimension of reality and its mechanistic nature and are thus limited in this sense.
Thirdly, it’s not like only free will is difficult to make sense of in naturalism. Philosophers have found out that something as basic as a “belief” is very difficult to fit within a naturalistic reality, i.e. to describe what exactly a belief is within a naturalistic reality. It is on such grounds that Alvin Plantinga proposes yet another argument against naturalism (see Plantinga’s recent book-length debate with non-theistic philosopher Michael Tooley in “Knowledge of God”). My point here is that not only the “freely choosing element” of our nature is difficult to make sense of within naturalism, but even much more basic elements of the human condition (such as having beliefs), elements which nobody dares call “illusions”, are equally difficult to make sense of within naturalism, at least by those who think carefully about such matters.
[continued in the next post]
[continued from the previous post]
ReplyDeleteNow back to the issue of personal free will. We do agree that all other personal elements or “forces” (such as beliefs, desires, feelings, virtues/character, states of affairs, etc) are not sufficient to account for the reality of free will. Some other element is there, the freely choosing element of our nature, which makes the final call once the probabilities are set by the pulling and pushing of the other forces. Now you ask “on what does that [freely choosing element] make its call?” The answer is, “on nothing”. That’s why that freely choosing element of our personal nature is sovereign and creative: It transcends all other elements of our personal nature by being an (or perhaps “the”) *uncaused cause* of our nature. Given the strength of the other personal forces, the freedom making capacity of that element is limited, but its cumulative effect is significant (think of the “sailor” analogy). Further it seems that if one steers one’s life towards the “good” or “natural” direction, the strength of the freely choosing element grows and we become “masters” of ourselves, a goal which all great spiritual teachers and mystics in human history have greatly valued. That is a state in which the other personal forces work harmoniously with each other. This is, I believe, the state which in Christianity is referred to by the prayer “Your will be done, not mine”. For God has created us in His/Her image, and thus our will, when we fulfill our design plan and become similar to God, is one which easily and naturally overlaps God’s, for in that state our beliefs, our values/desires, our feelings, our virtue/character, are all similar to God’s.
So far then my account about free will in a theistic reality. We are free to steer our personal reality far from God, but in doing this we remain on choppy seas. It is by steering towards God that we find the tranquil waters and the joy of personal fulfillment. I described this matter in some detail to preempt the following possible question: “If the freely choosing element of our nature is sovereign then perhaps one should not say that the choosing itself is random, but it would seem that the path one chooses is random.” To use the sailor analogy, the sailor may freely steer the boat in any random direction. On the theistic worldview not all directions are the same though; rather reality is such that one’s choices have implications, namely they drive one towards the calm waters where one feels at home, or else they drive one towards the choppy seas where one feels torn. Any which way one chooses to steer one’s life one will not fail to experience that difference, a difference between goodness and evil, and which will move the rational person towards the right decision. (Incidentally, the realization of that moral structure of our life may or may not go hand in hand with knowledge of God.) There is an implicit moral order in our experience of life, an order which we all can notice and make good use for, even when we don’t realize that that order points towards God. It is in this additional sense that ethics and freedom are connected: one not only needs freedom for ethics, but also ethics for freedom.
[continued in the next post]
[continued from the previous post]
ReplyDeleteNow, here is why I didn’t see the problem before: On theism the fundamental structure of reality is personal, and thus our own personal nature is in its structure irreducible to anything more basic. Freedom of will then, as we experience it, is a given part of fundamental reality and can only be described but not analyzed or reduced to other facts. It is the other facts that are contingent and have to be accounted for on personal terms. Thus, given my experience of free will, it never occurred to be to ask “on what do I freely choose?”, never mind to ask “does who is choosing actually exist?” My experience of free will is that, even though it is very strongly influenced by other forces, the final call is sovereign and mine alone. To the degree we experience freedom (and we do experience that freedom almost every second of our life), choosing is something one oneself *does* and thus is nothing one may compare to “randomness”. It is a fact of our condition that it is us who choose and that our choosing is *not* like throwing dice. (Which reminds me of something the German philosopher Schopenhauer once said: “materialism is the philosophy of the subject who forgets to take account of himself".)
Conversely, on naturalism the fundamental structure of reality is mechanical, and thus all that happens must follow mechanical laws and must either be determined by something else, or, to the degree it is not determined, must be random. Therefore there is no place for free will as we experience it in a naturalistic reality. What I could not understand is why non-naturalists (such as agnostics) and even theists could find free will to be a difficult, even “incoherent”, concept. Now I see that perhaps the reason is that such agnostics and theists, while not subscribing to naturalism, are nevertheless used to thinking about reality in naturalistic/mechanistic terms, namely as a matrix of lawful causes and effects in which conscious experience somehow manages to incorporate itself. Perhaps, being scientifically knowledgeable, they believe that there is no other way to conceptualize reality, for unless something is mechanizable, in the sense of amenable to describing within some kind of cause and effect paradigm, one is left with chaotic arbitrariness. But that’s the illusion, for in fact next to nothing in the human condition is thus mechanizable. Reality, being personal, is ultimately *not* mechanizable. Moreover, the way we are accustomed to speak may confuse the matter. For example I too speak of our “physical environment”, as if the physical realm was the big space out there in which we small spirits make a living. But on theism it is not the spiritual that exists within the physical (albeit with the spirit of God calling all the shots), but the physical that exists within the spiritual. Burk goes as far as to hold that it is irrational to think that one is not “co-extensive” with one’s body, as if there were a good reason to think that we have a physical size, and moreover one not greater than our body.
In conclusion, I say that the right way to think about a theistic reality is from the spiritual to the physical and not the other way around. We should take our own personal reality as the conceptual basis by which to understand everything else. This epistemological principle, I say, is implied by the very thesis of theism according to which all reality is God-grounded and therefore ultimately of a personal nature. Conversely, if the theist commits the epistemic error of thinking about reality from the outside in, and tries to conceptualize personhood in terms of non-personhood, then the theist is apt to get about as confused as the naturalist who tries to understand theism on naturalistic terms.
JP
ReplyDeleteThat's very interesting, I had a similar initial reaction to Dennett. As I understand it, he makes three key points on free will.
First, the traditional, intuitive concept of free will doesn't work very well when we pull it apart. Specifically, the only way to defeat the Laplacean demon is to introduce randomness which adds nothing valuable to our sense of morality, purpose etc. I agree with this.
Second, evolution has created what he calls evitability, the capacity to adjust one's behaviour to match our expectations for the environment, along with a level of self awareness which allows us to construct narratives about our behaviours. This allows a free will of sorts to occur, albeit one that is a little different from our intuitive sense of it. I agree with this two.
Third, this new perspective on free will does not cost us anything. There is nothing lost in this conception and this is a 'freedom worth having'. On this last point I remain uncertain. I can not convince myself that things like the very definition of belief, as Dianelos has mentioned, might not be threatened by this stance.
I can't get my head around this point yet, but as I've already mentioned, if our very concept of reason balances upon some notion of free will that is at odds with Dennett's construction, then the whole thing would fall apart. Something for me to think about.
Bernard
Dianelos
ReplyDeleteThank you for your thoughtful response. I am coming closer to understanding your point of view.
As always, the way we use language is tangled up in this thing. So, when you say the final decision made by the self is based on nothing, that comes very close to what I would call random. Specifically I struggle to see how a decision based upon nothing would add anything at all to my sense of self or stewardship. The coin tossing captain and the captain setting course based upon nothing at all feel like very similar sailors to me.
Bernard
Hi Bernard,
ReplyDeleteYou write: “As always, the way we use language is tangled up in this thing.”
Right, but my purpose here was to *describe* my experience of freedom, which I trust is the same with everybody’s else. Further, as I don’t know of the slightest evidence against the reality of my free will, I have no reason to doubt in the reality of it. But if free will is real, then we have strong evidence against naturalism, because, as I think we agree, free will does not fit within a naturalistic reality. Conversely, as I tried to describe, free will does fit very well within a theistic reality.
“So, when you say the final decision made by the self is based on nothing, that comes very close to what I would call random.”
I meant “based on nothing else but me”. Far from being random, the *final* decision is determined, indeed determined by me, and indeed by me alone and *nothing* else. And that’s why, to the greater or lesser extent I am free to choose, I am fully personally responsible for my choices.
Hi, Dianelos-
ReplyDeleteYou sound a bit like the person who insists that the sun comes up, therefore the sun goes around the earth. While we might call the geocentric model an "illusion", a better way to describe it is an impression we get from partial information (the "experience"). With more information, we get a better view of how the sun and earth relate.
Likewise, the "me" that you insist is free, that everyone else knows is free, responsible, but not random.. what is it? I think we all agree on the sensation of free will. Not knowing what we are going to do next, our decision to eat that candy bar seems free. The issue is what the details are. If I eat because I am hungry, was that free? If because I am nervous, was that free? If because my transcendent self takes pity on my bodily self, was that free?
If there is a reason, then the decision wasn't free, and if there isn't a reason, then what kind of freedom is that?
Dianelos
ReplyDeleteI don't yet understand what you mean when you say a final decision is 'determined by me and indeed by me alone.' If this determination must float free of all the salient factors that define the probability function (so it is made without regard to moral values, circumstances, expectations, personality, knowledge of God...) then that is I think identical to a random process.
What I mean by this is that if we compare self determinations to random determinations, we will not be able to find a pattern of difference in the outcomes. (If we could then we could presumably start the business of hunting down the correlates of these differences, and these correlates could reasonably be thought of as external determinants, so are part of the probability function).
Bernard
Hi Burk,
ReplyDeleteYou write: “You sound a bit like the person who insists that the sun comes up, therefore the sun goes around the earth.”
I agree that almost all beliefs we hold can be wrong. On the other hand, one should accept those beliefs that strike one as obviously true, unless very good evidence against them is forthcoming (a so-called “defeater”).
Now the belief that we are free to choose is as close to obvious as it gets. For example I have much more confidence in the existence of my free will than in the existence of other minds. Nevertheless I may be mistaken. So where’s the evidence against my belief? After all there is hardly any claim more extraordinary than the claim that we are not really free; so where’s the extraordinary evidence for that claim?
To my knowledge there is not one scintilla of evidence against the reality of personal freedom. All arguments I know are cases of begging the question, for they do little more than point out that one cannot describe free will in naturalistic terms. But this only evidences that free will and naturalism are incompatible, or that one cannot think about free will using impersonal/naturalistic concepts. Similarly one cannot think about, say, qualia (such as “redness”) using impersonal/naturalistic concepts, but this of course does not mean that qualia do not exist. Rather it means that the existence of qualia represents one more conceptual problem for naturalism.
Moreover, if free will does not exist then we are not responsible for our choices, which implication is as absurd as it gets. Given then that 1) it is prima facie obviously true that free will exists, 2) there is no evidence whatsoever against the existence of free will, and 3) the premise that free will does not exist implies absurdities – it is in my judgment irrational to believe that free will does not exist. Actually, given this epistemic state of affairs, it is in my judgment irrational to even entertain any doubts about the existence of free will. And if metaphysical naturalism is incompatible with free will, then so much the worse for naturalism. I mean it’s not like there is other good evidence for naturalism anyway.
“Likewise, the "me" that you insist is free, that everyone else knows is free, responsible, but not random.. what is it?”
Surely you are not asking me to describe the freely choosing element of our personal nature in terms of molecules, or of physical forces.
“If I eat because I am hungry, was that free? If because I am nervous, was that free? If because my transcendent self takes pity on my bodily self, was that free?”
As I think the previous discussion with Bernard makes evident, forces such as feeling hungry explain how come there is a probability distribution among the various options open to one, but do not explain the final call. It is in that final call, a call within the wiggle room left open by the probability distribution caused by all the other personal forces (feelings, beliefs, desires, etc), where the power of our sovereign freedom lies.
“If there is a reason, then the decision wasn't free, and if there isn't a reason, then what kind of freedom is that?”
The only kind of freedom there is: The power to transcend the implications caused by one’s current personal state, as given by the sum total of the other personal forces such as feelings, beliefs, desires, etc. Again, I am not here suggesting a theory of free will, I am only trying to describe to the best of my ability what is the freedom we experience. Further my argument is not that our experience of free will is coherent, because, trivially, all experience is coherent by definition. My argument rather is that our experience of free will 1) does not in any way contradict science, and 2) coheres perfectly well with the theistic worldview. Indeed our experience of free will is of fundamental importance for the theistic worldview, for it is in our freedom that we, perhaps most conspicuously, are made in the image of God.
Hi Bernard,
ReplyDeleteYou write: “What I mean by this is that if we compare self determinations to random determinations, we will not be able to find a pattern of difference in the outcomes.”
I agree. Indeed if we were to examine a human body objectively we would not find any evidence that the “final call”, which I claim is caused by our transcendental free self, is the result of anything else but a random process. In other words based on the scientific method we do not find any reason to believe that freedom of will exists. But why do you think this fact has any relevance whatsoever to the discussion at hand?
There is an idea, called “scientism”, according to which the only appropriate way to discover truths is the scientific method. To my knowledge scientism is not anymore defended in serious philosophical discourse, but it is certainly advertised in popular books such as Daniel Dennett’s. Scientism is a very strange belief for which no justification is given, beside to point out the success of the scientific method, as if the success of the scientific method in the scientific field somehow implies that the same method must be successful in other cognitive fields also, and, even more arbitrarily, that no other method can be successful in any other cognitive field. Now if scientific naturalism were true, i.e. if it were true that reality consists only of the kind of existents that science posits in order to model physical phenomena, then scientism would be true also. But to implicitly assume scientific naturalism is, once again, to beg the question. On the contrary one can easily enough see that scientism is false. After all, by scientifically examining the human body we don’t find any evidence for the existence of qualia either (we don’t find “a pattern of differences in the outcomes” as you put it), but surely this does not imply that qualia do not exist. (Actually one can point to the failure of scientism as evidence against scientific naturalism.)
Hi Dianelos
ReplyDeleteYou're cheating here, by equating reason with scientific process. And, as your arguments appeal at length to your own conception of reason, this is a little unfair.
I am arguing simply that if there is no observable difference between random and willed decisions, then this has implications for moral responsibility for instance. Not a whiff of scientism here.
The statement 'I make free decisions' when viewed through the filter of 'and these decisions are consistently random in their outcomes' does look rather different. That's all.
You are quite right to observe that one can always retreat to a position of pure faith (it looks like randomness but it just isn't because I know it isn't so there) but it's not helpful to engage in a conversation on these grounds, as such fundamentalism is unavailable to social scrutiny.
Bernard
Hi, Bernard, Dianelos-
ReplyDeleteForgive me for commenting again, but this is very interesting. I take Dianelos's accusation of scientism as meaning that he is unwilling to look beneath the hood of "free will". He claims it is obvious, self-evident, and that is that. Scientism is the insistence that everything is composed of parts, and the more we understand of the parts, the more we understand of the whole. So we are taking a reductionistic approach to free will, which is summarily rejected by the theist.
This applies across the board to many other topics, so it seems a very broad theme in this debate- whether at some point we can preserve our subjective experience from reductionistic intrusion and, ultimately, vitiation. It is a very live issue indeed.
Hi Bernard,
ReplyDeleteYou write: “I am arguing simply that if there is no observable difference between random and willed decisions, then this has implications for moral responsibility for instance. Not a whiff of scientism here.”
So let’s forget about scientism. I thought you embraced scientism, but obviously I misinterpreted your stance.
Now I agree that there is no observable difference between random and willed decisions (within the probability distribution caused by the plethora of motives that affect a choice). So what do you think is the relevance of this fact? You say that this fact has implications for moral responsibility, but I don’t see how. If in fact our choices are willed then there is moral responsibility, but if alternatively our choices are random then there is no moral responsibility. It seems then that the fact that no observable difference exists tells us nothing about whether moral responsibility exists or not. (This only shows that there are cases where objective observations are not sufficient to decide one way or the other. There are many other such cases.) But if the fact that there is no observable difference between willed and random decisions, then that fact is irrelevant in our discussion.
I also would like to have your take on my epistemic stance:
1. It strikes me as obvious that my decisions (my “final calls”) are determined by me and are nothing like random. Therefore I believe in principle in the existence of free will.
2. I know of no evidence whatsoever that would move me to doubt the existence of free will (i.e. I know of no potential “defeater” for my belief in free will). Indeed the existence of free will is compatible with all observations, scientific, objective, or subjective I know of. Thus I am quite confident that free will exists.
3. On the hypothesis that no free will exists and that my decisions are ultimately random there is no personal responsibility. I judge this implication to be absurd, which moves me to reject that hypothesis and strengthens even more my confidence in the existence of free will.
So I believe in the existence of free will, because I find there are good reasons to believe it, no good reasons to doubt it, and good reasons to disbelieve the opposite. Few beliefs I hold are as well justified as this one. Do you see any error in my justification?
Hi Dianelos
ReplyDeleteI think there might be a number of errors here. The first is the statement that your decisions are nothing like random. Well, in terms of outcomes, they are exactly like a random process. They produce, by your own admission, decisions that can not be distinguished in any way from those generated by some randomness generator.
That's an error of overstatement I suppose. The next may be an error of reasoning. If your final calls appear random from the outside, then they are constrained to appear random. They must appear so, and this would at first blush appear to limit the type of freedom you have.
Finally, an error of implication. Your freedom, thus diminished (it is stripped of its ability to affect the determined probability function, or to follow patterns of will that would make it distinguishable from randomly generated decisions) may not imply moral responsibility. I would argue that moral responsibility stems not from the way the probability function is resolved (in fact my own suspicion is that there is nothing for the randomness to do, in fact thought processes will turn out to be largely deterministic) but rather the way the distribution is built.
In essence, I think moral culpability is more likely to emerge from the fact that the stories we construct as individuals and societies set the parameters for our actions, than it is to emerge from a final probability collapsing function that appears entirely random, even though it feels otherwise.
As an aside, my own decision making process doesn't feel at all like the one you describe.
Bernard
Hi Burk,
ReplyDeleteYou write: “He claims [free will] is obvious, self-evident, and that is that.”
Not exactly. I claim that given that there is no evidence I know of against free will, and given the sheer presence of free will in my experience of life, the reasonable thing is to believe it exists. Especially given the fact that the alternative leads to absurdities, such as that we don’t have moral responsibility for our choices, for these are the result of a mix of deterministic and random effects, as is the mechanical output of any real world machine.
“Scientism is the insistence that everything is composed of parts, and the more we understand of the parts, the more we understand of the whole.”
I don’t think that this is what scientism is about, or is even implied by scientism. After all the idea that “everything is composed of parts” leads to an infinite regression of the type “it’s turtles all the way down”, a view which I think virtually nobody finds reasonable.
“So we are taking a reductionistic approach to free will, which is summarily rejected by the theist.”
Actually it's the other way around. The naturalist holds that the “final call” in our process of choosing is entirely random, and thus irreducible by definition. Thus the naturalist’s position is that there is nothing under the hood. On the contrary the theist does hold that there is an engine under the hood, namely the sovereign transcendental agent who determines the final call.
(To be precise we shouldn't be speaking of naturalists and theists; the disagreement is between the so-called libertarian and compatibilist interpretation of free will, and there are theists and naturalists in both camps.)
Dianelos
ReplyDeleteYour model has got me thinking. The more I consider it, the more I struggle with the idea of wilfully collapsing a probability function. I'm not sure I'll be able to explain this, but let me try.
In a couple of weeks I will be voting in a local election, and at the moment I am divided between two candidates. One represents a leftist Green party, the other is an anti-poverty activist. Neither has a chance of winning. I think I have a clear picture of the various influences tugging at my sympathies and prejudices, but as yet this has not resolved.
Your model seems to suggest the decision will happen this way: my personality, history, knowledge etc will lead to a weighted predisposition. (So let's say these deterministic factors make me three times more likely to vote Green). But the deterministic algorithm is insufficient to yield a decision, it can only create probabilities. Now the self steps in, and freely chooses according to these weightings. This free choice, while wilful, does not take into account any salient features of the choice (who the candidate is, what they stand for, their chance of winning etc) as these are already built into the probabilities. So, the self isn't choosing between Green and union at all. Rather they are in essence choosing a random number (between 1 and 8 would do it, with 1 through to 6 equalling a vote for Greens).
But look at how very limited such a version of freedom is. Not only is this final wilful nudge unaffected by any of the factors underpinning my political leanings, I'd propose it can't even know what its decision for a number translates to (because if it does it reduces to choosing between two options and makes a mockery of the weightings).
So the self of your model may have agency, and may be free to 'choose', but it is forced to choose without reason or information. I don't think any sort of moral responsibility can be drawn from this process. Rather, a deterministic mind where the salient information yields the decision provides a more responsible voter, to my mind.
Perhaps I have misunderstood your model, but to use Burk's phrase, this is what I think we find when we peer under its hood.
Bernard
Hi Bernard,
ReplyDeleteYou write: “The first [error] is the statement that your decisions are nothing like random. Well, in terms of outcomes, they are exactly like a random process.”
But what one observes about the outcomes is irrelevant, because whether the final call is willed or else is random will produce exactly the same outcome.
Further, in the previous post, when I said that free will is nothing like random, I am using “is like” to express our experience of free will. And I trust you agree that our experience of free will, or rather of us making the final call once all other motives or forces have played out their influences on our process of choosing, is nothing like random. But if you are unsure about this, I invite you to make the following experiment: When confronted with a choice among two alternatives where the probabilities set by the other motives are about 50% for each, instead of making the final call yourself simply flip a coin. We both agree that the outcome will be the same, i.e. from the outside one will not observe any difference between Bernard not flipping the coin and Bernard flipping the coin. But if you try this I think you will realize that from the inside the experience will be very different: choosing what the flip of the coin determines is not at all experienced as being similar to what one normally experiences. (I have actually made that experiment long ago but for other reasons.)
“If your final calls appear random from the outside, then they are constrained to appear random. They must appear so, and this would at first blush appear to limit the type of freedom you have.”
Why? I don’t see how this follows at all. There are many cases where how things appear from the outside says nothing about how things actually are. So, for example, whether a being is conscious or not makes no difference to one’s observations of it. That’s why nobody knows whether cockroaches or thermostats are conscious or not. Indeed if reality were so that human consciousness depends on some physiological quirk that obtains in only 10% of the human brains, we wouldn’t notice any difference either. And even if one assumes that all humans are conscious, our observations cannot decide whether all people experience colors basically in the same way or not (see the so-called inverted spectrum paradox). There is thus a plethora of examples where how things are observed to be from the outside is insufficient to warrant beliefs about how things actually are.
[followed in the next post]
[followed from the previous post]
ReplyDelete“Your freedom, thus diminished (it is stripped of its ability to affect the determined probability function, or to follow patterns of will that would make it distinguishable from randomly generated decisions) may not imply moral responsibility.”
It is true that our freedom is influenced by the state of how we are (i.e. by all the other motives, and how they pull and push determining the probability distribution). But how we make the final call in turn influences our future state, and thus influences the probability distribution that will obtain in the future. It is in this sense one is responsible for one’s actions (and for how one is). I’ll agree with you that if one delimits just one particular choice then moral responsibility does not obtain even at the presence of free will. But real life consists of a huge number of consecutive choices.
Here is an interesting point that just occurred to me: We all instinctively understand that for moral responsibility to obtain a sufficiently long period of choosing and personal state formation (or “soul formation” in theistic-speak) is needed, and thus hold that children are not morally responsible. Further, for moral responsibility to obtain we understand that the natural interplay between freedom and “soul formation” must be in good working order, and thus hold that mentally ill people are not morally responsible either. Significantly, both these broadly accepted principles about moral responsibility are directly explained by my account of free will. Conversely, it seems to me, they are not explained on the compatibilist account. Here then we have one more argument for the existence of free will, this time based on abductive reasoning.
“In essence, I think moral culpability is more likely to emerge from the fact that the stories we construct as individuals and societies set the parameters for our actions, than it is to emerge from a final probability collapsing function that appears entirely random, even though it feels otherwise.”
Hmm, if I understand you correctly then I’d like to point out the stories we construct are themselves the result of our past choices, and if free will does not exist one does not have moral responsibility for them, and nor therefore for their influences on the probability distribution. I think there is a vicious circle here: if there is not a sovereign/motives-transcending/ creative agent who makes the final call there is no way for moral responsibility to take. But if there is such an agent then it is transcendental, precisely because its existence makes no observational difference. That’s why, it seems to me, moral responsibility only makes sense in a religious understanding of reality.
Hi Dianelos
ReplyDeleteI'll wait for your response to the voting dilemma before giving a full response, as the more I think about it, the more I believe your notion of free will is ill defined, and I'm interested to see if you can explain how this wilful collapsing of probabilities could possibly work.
As you know, your consciousness example doesn't work for me. Like free will, I suspect the notion is poorly defined, and indeed consciousness is as consciousness does. If we ever understand the mechanism of consciousness properly, we may be able to infer its existence purely by function.
Your assumption that moral responsibility requires free will is an assumption with its own built in conclusions. I'm not sure I buy it.
Bernard
Hi Bernard,
ReplyDeleteYou write: “This free choice, while wilful, does not take into account any salient features of the choice (who the candidate is, what they stand for, their chance of winning etc) as these are already built into the probabilities. So, the self isn't choosing between Green and union at all. Rather they are in essence choosing a random number (between 1 and 8 would do it, with 1 through to 6 equalling a vote for Greens).”
What I tried to do is to *describe* our experience of choosing, and the above is not a think a good description of how we experience making the final call, in which the whole process of choosing comes to a closure. Rather, after having pondered the issue and having allowed all other motives to play themselves out, there comes a moment where we feel that now is the time to make a stand, to express our will and affect the flow of things. And it’s not like when one determines the final call one does not take into account the salient features of the various choices but simply mentally chooses a random number. On the contrary it is now that one pulls all the threads together in order to be aware of one’s various motives for or against the various choices, and while facing them makes the sovereign and creative determination that expresses one’s free will.
At this juncture I’d like to make two points: First of all I am aware that the way I have been describing our experience of choosing is only a first approximation. In real life we make “final calls” all the time, including while thinking, say, about whom to vote for in the next election. The reality of our experience of freedom is more dynamic than what I have been describing, but I think my description does capture the salient features of that experience to a degree that allows one to see how that reality fits in the large scheme of things. Secondly, it seems to me that nobody is really suggesting that we do not experience life as being free in the first place. That’s why many naturalists suggest that our experience of freedom is an illusion, in the sense that we are not really free to choose even if it feels this way. Thus, any realistic description of our experience must be one that comports with freedom.
“But look at how very limited such a version of freedom is.”
Freedom is a binary property; either it is there or it isn’t. Perhaps in one particular state of affairs the space in which we can determine a free choice is very limited indeed, but 1) it is still freedom, and 2) the cumulative effect of many small free choices can become very large. Incidentally, something similar goes for consciousness: Either a being is capable of experiencing or it isn’t. If it is, perhaps it is capable of experiencing only very little, but it is still a fully conscious being.
“So the self of your model may have agency, and may be free to 'choose', but it is forced to choose without reason or information.”
But that’s the opposite of the experience I was trying to describe. One determines the final call by taking into account all types of reasons/motives which determine the probabilities of how one will choose. On the contrary, what I think will feel like “being forced to choose without reason” is to toss a coin and really let random chance play the role that one normally plays when one makes the final call. That’s why I have been saying all along that our experience of free will is nothing like random.
Perhaps the difficulty you are facing is to try to rationalize experience from the outside, but experience is a brute given. The only thing one can reason about is what one’s experiences imply about the rest of reality.
Dianelos
ReplyDeleteWhile your experience may be a brute given, your interpretation of it is not, and this is where philosophy comes in.
At this point you are running a contradiction in your interpretation of what you take to be the brute fact. I am happy to accept that your experience of freedom does not feel at all like the process I have outlined. However, your description of this freedom implies the process I have outlined. And this means that somewhere between your experience and your description, a misstep is occurring. Only you have your infallible experience of being free, so you'll have to work out what the misstep is, but the contradiction is there front and centre I'm afraid.
The contradiction, in case you've missed it, is this. You speak of your free decision taking into account all of the factors that underpin it, and yet you also speak of your free decision causing the final call once all of the deterministic conditions (taste, mood, sobriety etc) have been accounted for. So either your final call is blind to the conditions (free) or takes them into account (determined). Unless we choose to step outside the ordinary rules of logic, one can't have both.
You may choose to say, as Eric does, that while you can see no way through the logical impasse, you give preference to your intuition over logic in this case. I think that's an honest and admirable admission, and it may be we all have to do this upon occasion to knit our various impulses into meaning.
Bernard
Hi Bernard,
ReplyDeleteMoral responsibility seems to be at the core of the whole issue – and free will simply a way to get there. I must say I find the whole thing troublesome and I will try to sketch what troubles me.
First, moral responsibility does not seem to follow at all from free will (of any kind). All, I think, will admit that our decisions are largely the result of our background (knowledge, family, life experience, social context, and so on). As we are not responsible for most of these factors, how can we be morally responsible of actions deriving from them?
Second, and although this may sound like a strange question it is a serious one, what is moral responsibility good for? Of course, in religious settings with a notion of salvation or some similar reward and punishment scheme, this may be the way to go. But, outside of this, why do we need it at all? We can have a perfectly sound justice system designed to facilitate an harmonious social life, based maybe on a notion of “practical” responsibility, allowing for internment of dangerous persons and deterrence measures. Introducing a notion of “good” and “bad” does not appear to add anything of value.
Lastly, maybe the idea moral responsibility is not necessary for social life, but it is innocuous? I can see two (related) ways in which it can be harmful. Once a person has been found morally (and legally) guilty and “punished” for it, it is easy to conclude he has only what he deserves and that, consequently, we have no longer any obligation to treat him as a full human being, to care for him except in the most basic ways. We put dangerous animals in cages to protect ourself but they are, I would suggest, better off than many humans in our prisons.
Alternatively, and this may be very pernicious I think, when we assign the blame for a crime on a single person what we do at the same time is make all the rest of us innocent - one goes with the other. This is of course a very convenient doctrine as it removes the need to acknowledge complex social causes and, indeed, our own personal responsibility as part of a huge social machinery causing a lot of injustices to many while bringing huge benefits to others.
Hi JP
ReplyDeleteI suppose I'd need to know what you see as the distinction between moral responsibility, which you're in some way arguing against here, and personal responsibility which you refer to in order to construct that argument.
I absolutely agree with you that the most productive thing we can do in the face of any form of social breakdown is examine our own contribution to it. But doesn't that show precisely how important the notion of agent responsibility is?
Bernard
Hi Bernard,
ReplyDeleteYou write: “While your experience may be a brute given, your interpretation of it is not, and this is where philosophy comes in.”
I completely agree. And it is in the *interpretation* of an experience where potentially the illusion comes in. So, strictly speaking, an experience can never be illusory because the experience is exactly as it is. Only how one interprets an experience can be illusory. A few common examples: “When we put a pencil in a glass of water it looks like broken, but this is an illusion for in reality it is not.” When we say this we don’t question the fact or the truth or coherence of the experience itself, for we do experience the sight of a broken pencil. The illusion here consists in misinterpreting the experience and believing that if one would take the pencil out of the water it would still look broken. Another example: “The asphalt at the end of the road in a hot day looks like wet, but this is an illusion for in reality it is not”. The illusion here is to misinterpret the experience and believe that when on reaches that spot of the asphalt and touches it one’s fingers will get wet. In think in all cases where we speak of illusions, we don’t mean the experience but a false interpretation of the experience, indeed an interpretation that at least in principle can be empirically falsified.
In the case of freedom some people claim this: “When we choose it looks like (at least up to a point) our choice is free, but this is an illusion for in reality it is not.” So here too, it’s not the experience itself that is deemed illusory, but one’s interpretation of it. In much of what I have been writing before, I was simply *describing* our experience of freedom, as it in fact is. As for the claim that one’s interpretation of the experience of freedom is illusory, I have been arguing that not a scintilla of evidence for that belief has been forthcoming. I now see that I could have made a stronger claim, namely to point out that the claim that freedom is illusory is not empirically falsifiable, as all claims about illusions should be. After all there is no difference between the observations made in a world in which freedom is real and a world in which freedom is not real. Thus it seems that the claim “our experience of freedom is an illusion” is a case of special pleading, because it does not conform to the standards of all other claims about illusions.
“However, your description of this freedom implies the process I have outlined.”
No, the process you have outlined is *compatible* with my description of the experience of freedom, but is not *implied* by it. Suppose we agree that A is true. That B is logically compatible or consistent with A does not imply that B must be true also. To think otherwise is to commit the so-called fallacy of affirming the consequent. (In our case A would be our experience of freedom, and B the process you have outlined.) By showing that your process is compatible with the given experience, you have only shown that your process *may* be true and nothing more. Or, to be more precise, that our experience of freedom does not falsify the process you outlined.
I would say that our experience of freedom is consistent with two processes: First, that the final call is random and not really determined by anything. Or, alternatively, that a transcendental agent determines the final call. How are we to choose which of these processes is the more probably true? Science does not help, because all scientific observations are compatible with both processes, so to reason about this we need philosophy.
[continued in the next post]
[continued from the previous post]
ReplyDelete“The contradiction, in case you've missed it, is this. You speak of your free decision taking into account all of the factors that underpin it, and yet you also speak of your free decision causing the final call once all of the deterministic conditions (taste, mood, sobriety etc) have been accounted for. So either your final call is blind to the conditions (free) or takes them into account (determined). Unless we choose to step outside the ordinary rules of logic, one can't have both.”
I don’t see that at all. Let’s call my interpretation of the experience of free will “agent causality”, and let’s use the example you suggested previously. Your various personal conditions (or motives or reasons) determine that the probability that your final call will be to vote for the Green candidate is about 75% versus 25% for the union candidate. We agree this far. The question is about the process that realizes the next critical step, namely the final call itself.
In the case of agent causality, once the probabilities are determined by the various conditions, a transcendental agent, while respecting the probability distribution and thus while being influenced by all the other conditions, determines the final call. So in agent causality the final call is *not* “blind to the conditions” and definitely “takes them into account”. Indeed, should the conditions be such that the probability of voting Green is 99.99% your final call will not only be influenced by the conditions but almost determined by them.
On the other hand, the conditions themselves do not determine the final call, for the final call is determined by the agent. In your actual example even though your specific conditions are such that it is relatively improbable (i.e. 25%) that you will vote for the union candidate, there is no force that can stop you from voting for the union candidate if you so determine. Even if the conditions were such that the probability of voting for the union candidate were much smaller, say 0.01%, there is no force that can stop you from voting for the union candidate, albeit it is quite improbable that you will actually determine to do that.
In conclusion, in this interpretation the transcendental agent, while being influenced by the conditions, makes a sovereign/creative/uncaused genuinely free final call. Also, contrary to the other viable interpretation according to which the final call is random (while also respecting the probabilistic distribution determined by the conditions), agent causality comports well with moral responsibility, and indeed explains some of its principle properties.
Coming back to what you say above then, in agent causality the final call *is* determined, indeed is determined by the transcendental agent. And, as you say, the conditions (the various motives, reasons, etc) are all taken into account, as they determine the probability distribution. So I really fail to see where the “contradiction” you speak of is.
Thinking about this it seems to me that agent causality is the simplest possible interpretation, for it says that the process by which the final call is made is exactly how making the final call feels like. (In this interpretation then the phenomenological and the noumenal are identical, thus solving Kant’s conundrum.) Thus, if experience, by its nature, cannot be incoherent or self-contradictory then it seems agent causality cannot be either.
Anyway, if after the above clarifications you still see a contradiction somewhere in agent causality then please point it out. I am certainly very much interested in understanding what contradiction you see.
Dinaelos
ReplyDeleteYes, the preamble is all a given, no arguments there. Your notion of agency carries a logical contradiction and this is, I think, very good ground for doubting it.
The final call is blind to the conditions, it is implied in your description of determining factors setting up a probability distribution. So, how can one choose between a 25%/75% split using a process that is not random, while respecting those odds?
Choosing a number at random from 1 to 4 does the trick, but how does your form of agency, which does not have access to such a random function, do it? I can see how your agent makes a choice, but not in a way that respects the parameters of the probability function without being random. That's where I think your contradiction lies.
Bernard
Bernard,
ReplyDeleteYou write: “The final call is blind to the conditions, it is implied in your description of determining factors setting up a probability distribution.”
How does a previous event in which various conditions determine a probability distribution *imply* that a subsequent event where an agent makes the final call according to this distribution “is blind” to the conditions which produced the distribution? I am really curious where you see that implication, because I don’t see any such connection whatsoever. If anything the implication would be the opposite: That the agent cannot possibly make a final call which respects the probability distribution while being unaware of it.
“So, how can one choose between a 25%/75% split using a process that is not random, while respecting those odds?”
The alternative interpretation of our experience of free will according to which the universe rolls the dice to choose among the 25/75 probability is logically viable. What I don’t understand is this: If a dumb rolling of the dice can achieve the end of respecting the 25/75 ratio, why can’t a rational agent achieve the same?
Or perhaps I am misreading your question. Perhaps you are asking “how” in the sense of “by which mechanism” does the agent achieve this end. If so, that’s a wrong question for nowhere have I claimed that the agent uses a mechanism in order to make the final call. On the contrary, agent causality is not reducible to any mechanical principles or order. After all, if the process that produces the final call were reducible to mechanical principles then it wouldn’t really be free, nor would there exist moral responsibility. So, if one perhaps unconsciously tries to project mechanical order onto a description of agent causality one will notice that it does not fit. But this of course does not expose an internal contradiction in agent causality, it only reveals that one is begging the question by assuming that everything must be mechanizable in the end.
Dianelos
ReplyDeleteI am perhaps explaining myself badly. Let me persevere. Feel free to say if you tire of this exchange, but I think we are getting very close to an understanding.
If I understand you correctly, you are saying there is a deterministic element to our decisions, but such determinism can not resolve a decision, it can only set up a probability distribution.
I think you are also arguing that all the things we might normally think of as determining a decision, so our tastes, knowledge, expectations, personality, culture etc, are part of this function.
So, I offered the example of an upcoming vote, where at the moment it feels sort of 75/25, were I to walk into the polling booth now that's what I'd need to resolve. Now, personally I believe were I to do this, there is a subconscious deterministic process that would kick in and lead to one vote feeling right, a sort of iterative process of reweightings leading to an eventual 100/0 split. But we're not discussing that.
We're discussing the model where there is alternatively a random component that collapses the probability distribution, or a free floating will that does the same job. I'm claiming the second option is contradictory. Perhaps a more moderate claim is that for it to work, the free will in question would need to be radically redefined.
Here's why. I walk into the polling booth. I have a 25% chance of choosing the union candidate, whatever this means, but then, on the spot, I decide this is what I shall do. So, on what grounds did I choose? Well, the answer can not involve any of my thoughts or feelings towards the two candidates, as all such knowledge is already built into the probability function. I can not do it because I feel a sudden overwhelming empathy with the poor, that predisposition towards empathy is how he registers at 25% in the first place.
This is what I mean by saying the free will in this case is blind. And at this point I find it very hard to see the difference between it and randomness. The collapse is caused by an agent that is in no way influenced by the pertinent factors.
Second, and you agree here I think, the pattern of willed decisions must over time appear wholly random. If not, then the factors beneath any pattern can be incorporated into the probability distributions. This means the wilful self can build up no narrative of its own, no set of feelings, characteristics or tendencies. The self has no accumulating history in other words, because this would violate the condition that its actions appear entirely random.
There may be a further objection to this notion of wilful collapse that relates to what we mean when we speak of probability in the first place, but in pursuing this line my enthusiasm will quickly outrun my expertise, so let's stick with these two points for now.
Thanks
Bernard
Hi, Dianelos and Bernard-
ReplyDelete"If so, that’s a wrong question for nowhere have I claimed that the agent uses a mechanism in order to make the final call. On the contrary, agent causality is not reducible to any mechanical principles or order."
I think we get to the nub with this assertion from Dianelos. He claims the agent/self is irreducible. This is reminscent of those who view biology generally as irreducible, by way of intelligent design and similar, ahem, theories. If one is unwilling to look at the composition of X, then X will look magical and in this case "free".
I won't go through chapter and verse on brain-psychology correlations, but the is no serious evidence that any portion of psychology comes from anywhere else.
Hi Bernard,
ReplyDeleteI'm not sure what is ordinarily understood by moral responsibility. This is a concept I find somewhat nebulous although obviously related to the notions of right and wrong (or good and evil). I assume we can speak of moral responsibility for an action when the agent or actor can be said to have acted willingly in accordance or against some moral code. And, if against, often follows blame, moral condemnation and “deserved” punishment.
It is with these latter aspects that I have some issues. This idea that someone “deserves” punishment strikes me as quite strange and I just can't make it work.
Maybe this is mostly a semantic issue, I don't know. I'm all ready to say, for example, that a student that does not work hard enough “deserves” bad grades. Or that, if I play lazily in a game of chess, I deserve to lose. But in these cases it does not seem to me that I am making moral judgments – they are entirely of the practical, causal kind. In making these judgments, I don't mean anything like the following: in not working hard enough, the student has committed a wrong, and deserves bad grades as “punishment”.
More generally I find that introducing moral responsibility may often lead to the perverse effects that I have tried to outline in my previous comment. My other point was that, in addition to having harmful effects, the concept of moral responsibility does not appear to be necessary to the implementation of a justice system. For example, fines and prison sentences, I think, may be entirely justified in terms of security and deterrence. I can't help seeing the idea of society punishing a person for moral failure as patronizing and childish. Moral responsibility is best left a private matter.
I admit I find articulating ideas on this topic very difficult but hopefully I am making some sense.
Hi Bernard,
ReplyDeleteYou write: “I walk into the polling booth. I have a 25% chance of choosing the union candidate, whatever this means, but then, on the spot, I decide this is what I shall do. So, on what grounds did I choose? Well, the answer can not involve any of my thoughts or feelings towards the two candidates, as all such knowledge is already built into the probability function.”
Right. I think I understand well what you mean. So, on what grounds does one choose? The answer, which I think I already gave before, is that at the moment of the final call (which is the culmination of the deciding process) the agent does not make that final call on some external grounds. Rather, the final call is a sovereign, creative, and uncaused expression of the agent’s own free will. At that moment the agent is not in any way, shape, or manner “blind” to the various motives that pulled and pushed the decision one way or the other, for the final call is greatly influenced by them. But the final call, to the degree it is free, transcends these motives – as I think you realize it must be the case. The various motives (which are an integral part of the process and indeed of one’s nature) influence but not fully determine the final call; the agent alone determines it by an act of its sovereign will. And that’s why personal responsibility enters the flow of the decisions we make in life, and why nobody can suggest that their life is like the path taken by a leaf floating on the rapids of a river, been tossed to and fro.
Perhaps the difficulty you feel lies in the following: You expect any effect (such as the agent’s final call) to have some sufficient reason or cause, or, if not, to be random. Which is OK. But then I suspect you also expect the sufficient reason or cause under discussion to be some kind of external force; something from the outside that moves the agent to make the final call. And, indeed, on agent causality there is no such external force; the force that causes/determines the final call is internal and integral to the agent and is called the agent’s free will. Now perhaps you have difficulty conceptualizing such a thing, but this does not mean there is a contradiction in the process I am describing. I myself have no problem conceptualizing sovereign free will, because that’s exactly how my making the final call feels like. So I am as familiar with sovereign free will as I am with the experience of redness.
“And at this point I find it very hard to see the difference between it and randomness.”
Seen from the *outside* there is no observational difference between free will and randomness being the grounds of the final call. I have already conceded that. And neither is there any observational difference between one having personal responsibility and one lacking it. Neither is there any observational difference between somebody being a conscious being and somebody being a zombie. Neither is there any difference between some people experiencing colors under an inverted spectrum. So what? Again, all these facts only evidence how pointless it is to try to understand reality using only concepts that make an observational difference.
[continued in the next post]
[continued from the previous post]
ReplyDelete“Second, and you agree here I think, the pattern of willed decisions must over time appear wholly random.”
Right. From the outside it is impossible to find out whether a final call or a sequence of final calls has been produced randomly or else willfully.
“This means the wilful self can build up no narrative of its own, no set of feelings, characteristics or tendencies.”
Why not? The process by which final calls are made may look random from the outside, but is nothing like random from the inside where the self lives and acts. Indeed, as I have described, the self’s power to guide its life is cumulative, and can be virtually absolute.
“The self has no accumulating history in other words, because this would violate the condition that its actions appear entirely random.”
The self, by determining the final calls, affects its life course and thus also the various motives or conditions (such as beliefs, desires, character, etc) that shape the probability function of future decisions. Thus, on agent causality, one’s current motives or conditions are shaped to a larger or smaller degree by the cumulative effect of one’s past free decisions. (The wiser one has chosen in the past the more master of oneself is one in the present.) So there is an accumulating history. From where I stand agent causality works beautifully well.
Or perhaps you mean that a process that is at bottom driven by randomness cannot possibly produce cumulative history, or something as complex as peoples’ lives. And therefore that a willful process that is indistinguishable from a process that is at bottom driven by randomness can’t either. If that’s what you mean, then your premise is clearly wrong, as the example of Darwinism evidences. Darwinian processes, even when unguided and ultimately driven by pure randomness, can and do produce impressive results.
Hi Burk,
ReplyDeleteYou write: “[Dianelos] claims the agent/self is irreducible. This is reminscent of those who view biology generally as irreducible, by way of intelligent design and similar, ahem, theories. If one is unwilling to look at the composition of X, then X will look magical and in this case "free".
Well, as we have been discussing, some things must be irreducible, otherwise it’s turtles all the way down. I suppose if I asked you by what process does mass bend spacetime, you’d answer that this is part of the irreducible nature of matter, right? And if I asked you how an electron manages to have electrical charge, you’d answer that that’s what an electron irreducibly is, namely an elementary particle that has electrical charge.
I assume we agree this far. Can you give me any other reason then why I can’t claim that the agent/self is irreducible in its free will, the way that mass is irreducible in its spacetime bending nature, and an electron is irreducible in its having electrical charge?
“I won't go through chapter and verse on brain-psychology correlations, but the is no serious evidence that any portion of psychology comes from anywhere else.
Are you saying that my hypothesis that one’s free decisions (the “final call”) are determined by a transcendental agent is contradicted by something we know about our brain? If so I’d very much like to know what that evidence is.
Or perhaps you are saying that I should not hypothesize a transcendental agent without evidence for its existence. But I do have evidence for its existence, namely that all other theories about how people decide imply absurdities such as that people could not have chosen differently from how they did, or that they are not responsible for their choices.
Hi, Dianelos-
ReplyDeleteI don't really share your view of the irreducibility of electron charge or mass-based graviation. We don't understand how all these things arise, that is true. And they are empirically connected in the most intimate way... that is also true. But whether they are irreducible or not is speculative. They are probably not reducible in any practical, hands-on way, but some string-y theory might come up with explanations for why these properties go together as they do.
"Are you saying that my hypothesis that one’s free decisions (the “final call”) are determined by a transcendental agent is contradicted by something we know about our brain?"
Absolutely. We know that all our decisions and thoughts come from the brain- that is the basic program & theory of today's cognitive scientists, neurobiologists, etc. The field wouldn't exist otherwise, and while we don't have positive knowledge of every thought and process, (yet), nothing that has been observed, from all the many techniques and observations available, contradicts the overarching hypothesis. Today, we are hearing so much about memory and cognitive deficits in former (American) football players- caused by lesions to their brains.
The original guest blogger himself offered some very cogent examples, such as that of Phineas Gage, whose lesions led directly to diametrically changed decision-making... what you claim is "free". In his case... not free at all.
You agree that the brain has something to do with free will and other thoughts, at least with transmitting them from wherever they originate, but seem to reserve something termed the "final call" as not originating in the brain, but rather from a transcendent source. Would Phineas Gage's "final calls" have been garbled by his lesioned brain, so that his transcendent self was unaffected, but these "final calls" couldn't get through and were substituted with some lesser brain-based decision that looked for all the world (and to him himself) as a final call, but didn't have the benefit of transcendent input?
Or conversely was his transcendent self likewise affected by the injury as his corporeal brain was, leading to its "final calls" being warped from the transcendent source? If the latter, then you would be proposing a theory, that, like vitalism and other theistic approaches to biology, simply grudges what biology finds while vainly trying to keep some "essence" or other ever-more inconsequential and udefineable element in the theistic camp.
If the former, you have given up the "final call" theory, since one can't really pick and choose who, by way of having a "normal" brain, has rights to the transcendent input, and who not.. you have no way to pick and choose, since here again, there is nothing empirical to go on other then the lesion that stares us in the face, from the brain.
Dianelos
ReplyDeleteHere's where I think you keep making a big leap. You say the circumstances influence but do not determine the final call. I say they can have no influence at all on the final call. They influence the setting up of the probabilities, not the way the probabilities collapse. And so, the final call, that aspect of the decision that chooses between deterministically unresolved options, is made by an agent with no regard to the factors that influenced the probabilities. The agent then is entirely blind to the circumstances of the decision (in that he/she can not take them into account), and hence is acting randomly. This is not, I don't think, a deficiency of my imagination, it's something you have yet to explain.
Bernard Beckett
Hi Bernard,
ReplyDeleteYou write: “Here's where I think you keep making a big leap. You say the circumstances influence but do not determine the final call. I say they can have no influence at all on the final call. They influence the setting up of the probabilities, not the way the probabilities collapse.”
It seems to me that you are conceptually separating the probability distribution (determined by the various circumstances) and the final call, or the “probabilities collapse” as you put it. But the probability distribution is not in any way independent from the final call; quite on the contrary the probability distribution is what *describes* the measure in which the various circumstances influence the agent’s final call. The very concept of the probability distribution depends on the agent being influenced by the various circumstances. If there weren’t any such influence then there wouldn’t be any probability distribution. You can imagine then how strange it sounds to me when you keep insisting that the circumstances do not influence the agent’s final call. In the process I am describing they certainly do.
“And so, the final call, that aspect of the decision that chooses between deterministically unresolved options, is made by an agent with no regard to the factors that influenced the probabilities.”
But as a matter of experiential fact, when we make a final call we are not aware of probabilities; rather we are aware and are being influenced by all circumstances that pull and push on our decision. So it is plainly and obviously false to say that we make the final call with no regard to circumstances such as beliefs, desires, etc.
I have the feeling that you are stuck with the process where randomness is what determines the final call. I understand that process, agree it is logically viable, agree that it fits with all observations – but this is *not* the process I am describing. The process of agent causality I am describing is one which takes at face value our experience of free will, i.e. one where the process of making a free decision is identical to how we experience that process. Please forget for a moment the random agency idea (it’s a pretty wild idea anyway, as it implies that our brain is massively fooling us in our experience of free will), forget any naturalistic or materialistic assumptions you may often use when thinking about consciousness, and just consider the process I am describing in its conceptual and logical merits.
Hi Dianelos,
ReplyDeleteI am also puzzled by your decision model. Here's what I understand.
There on one side an agent and on the other side all the various inputs relevant to a given decision (I include in these inputs the complete internal state of the agent). To simplify, let's say the decision process is a chain of steps in which the agent considers various inputs, arrives to an intermediate result that is then fed into the following steps, and so on.
At each step, the agent's action results of the consideration of inputs available at the time. If we replayed the tape, so to speak, the agent would have exactly the same inputs and would, presumably, take the same actions (leaving aside random elements). If so, then each step is determined and so is the whole chain.
If not then, at a given point in the chain, the agent would have to act differently on exactly the same inputs. The question is obvious: if the different outcome does not depend on a difference in input, what causes it? What else is there besides all these inputs? Remember that I included the complete internal state of the agent in the inputs.
You give particular importance on the final call, but it does not seem different from any other step.
Hi JP,
ReplyDeleteYou write: “The question is obvious: if the different outcome does not depend on a difference in input, what causes it? What else is there besides all these inputs?”
The agent’s freedom to make the final call. Agent freedom is irreducible and uncaused (otherwise it wouldn't be free, would it?) and that's why I have been describing it as "sovereign" and "creative".
Agent freedom is something we don’t find in the mechanistic models by which we think about physical reality (in which all causes are either deterministic or random), nor is agent freedom some specific part we employ when building machines. But this does not mean that agent freedom is “incoherent” or “contradictory” or a “big leap”. It simply does not fit with a mechanistic conception of reality, which, as far as I am concerned, is not a problem for the concept of free agency but a problem for the mechanistic conception of reality. For free agency is what we clearly experience almost every second of our lives, and unless there is extraordinarily powerful evidence against it, there is no reason whatsoever to doubt its reality. And I notice that there is no evidence against free agency whatsoever, not the slightest bit.
Hi Burk,
ReplyDeleteYou write: “But whether they are irreducible or not is speculative.”
But suppose they are. Suppose that naturalism is true and that it is an irreducible fact of nature that mass bends spacetime around it. Then it would be unreasonable for a theist to ask a naturalist about *how* mass bends spacetime, correct? All worldviews then posit some facts as irreducible, and theism posits facts pertaining to personal reality as irreducible (and explains all other facts, including physical facts, as reducible them). Thus it is unreasonable for a naturalist to ask *how*, say, the agent irreducibly determines the final call.
“We know that all our decisions and thoughts come from the brain- that is the basic program & theory of today's cognitive scientists, neurobiologists, etc. The field wouldn't exist otherwise,[snip]”
Let’s be clear on one thing: It is false that all these scientific fields would not exist unless consciousness is produced by our brain. Suppose we all exist in a computer simulation where our consciousness is not produced by our brain, but by a computer existing in a reality we have no inkling about. Even if that were the case the scientific fields you mention would proceed exactly as they do now, right? Therefore it is also false that we know that our consciousness is produced by our brain. We don’t know anything of the sort, and many philosophers have found that this belief is untenable. Incidentally, as far as I am concerned all the talk about neurobiology etc is as a red herring. Two thousand years ago everybody already knew that a strong rap on the head renders people unconscious, and that gouging their eyes out renders them blind. So it’s not like philosophers two thousand years ago did not know about the intimate relationship between consciousness and our physical body, nor like modern brain science has produced some dramatic new piece of evidence which was unknown then.
“Today, we are hearing so much about memory and cognitive deficits in former (American) football players- caused by lesions to their brains.”
Right. And if in a windowless room the light bulb is damaged then I will experience a deficit in my capacity of seeing. Until such time as the light bulb is fixed of course. Similarly, if the former football player’s brain is fixed they will get back their normal memory and cognitive capacities. Therefore, on this type of evidence it is as reasonable to conclude that consciousness is produced by the brain as it is to conclude that it is produced by the light bulb.
The only thing we really know is that, at least in our current condition, in order to have a normal conscious life our physical environment (which includes our brain) must be in an appropriately normal state too. Which says the obvious, namely that our physical environment is a delimiting factor to our experience of conscious life – but says nothing about where our consciousness itself, i.e. our capacity to experience a normal kind of life in a normal environment, comes from.
Hi Dianelos
ReplyDeleteI still think you're cheating. You describe a mechanistic process, that which determines the probability distribution, or if you will, the pull of various options. You then claim a sovereign power that transcends the deterministic factors, which to say is not determined by them. You say that despite the various pulls and weightings, the final call is still free to chose as it will. So you ask for the final call to sit outside the probability distribution. But then, when we explore the implications of this separation, namely the difficulty of a non random process providing this collapse, you say the final call is not independent of the probability distribution. Well, there's your contradiction isn't it, because technically speaking, if it's not independent of it, then it can't escape its determinism can it?
When I speak of the final call, I speak purely of the aspect that you say floats free of the deterministic influences. That part, by definition, is independent, that's what this floating free means. So, the problem for you remains. How can it be independent of these influences, yet not random? That's the question you appear unable to answer, and that causes a logical problem for your model.
Bernard
Hmm, my last post in this thread appears to have disappeared, so here's a shorter version of it:
ReplyDeleteHi Bernard,
You write: “You then claim a sovereign power that transcends the deterministic factors, which to say is not determined by them.”
No, I’ve been saying that the agent’s final call is not *fully* determined by such factors, and that it is precisely in that fact that the agent’s freedom rests. On the other hand in many cases the final call is so strongly influence by other factors that it is virtually determined by them. For example it is virtually determined that I will not decide to shoplift.
“You say that despite the various pulls and weightings, the final call is still free to chose as it will.”
On the contrary I have been saying, or have been trying to say, that the agent’s final call is strongly *influenced* by the various pulls and weightings, and that therefore the agent is *not* free to chose as it will. Further I explained that’s precisely what the probability distribution actually reflects. The probability distribution refers to or describes or quantifies the measure in which the various pulls and weightings influence the agent towards various alternative final calls, and thus it also describes or quantifies the measure of freedom still open to the agent. For example, if the pulls and weightings are such that the probability distribution describes a 99.99% probability of me paying rather than shoplifting, it also describes a very small measure of agent freedom I have in this matter. But some measure of freedom, large or small, is always available, precisely because the final call is never fully determined by the pulls and weightings.
“When I speak of the final call, I speak purely of the aspect that you say floats free of the deterministic influences.”
Ah, perhaps here lies the misunderstanding. When I speak of the “final call” I simply mean just that, namely the final decision, for example your final decision of whether to vote green or union. I do not mean “the aspect that floats free” of the various influences, as you presumed. What floats free of the various influences is the agent’s freedom to determine the final call, to the larger or lesser degree that this freedom exists in each case.
Dianelos
ReplyDeleteWell, this solves nothing at all for you I'm afraid. Look carefully at your last statement and you speak fo the agent's freedom to determine the final call. It's this freedom that floats free in your model, and hence faces the problem. I'm uncertain why you're avoiding this clear contradiction.
Bernard
Hi Bernard,
ReplyDeleteI must say I still don’t see any conceptual problem whatsoever for the concept of free will. On the contrary our discussion about what the problem might be has helped me sharpen my understanding.
I found an article in wikipedia about the “standard argument against free will” where many philosophers are quoted describing the problem (see http://en.wikipedia.org/wiki/Standard_argument_against_free_will). That argument appears powerful because it does not make any prior assumptions about reality, say about whether reality is theistic, or naturalistic, or demon haunted, or a computer simulation, or whatever. Nevertheless I think this argument in all its forms suffers from the same error: After rightly observing that to the degree a choice is not fully determined by the previous state of the world it must look random, it falsely assumes that what looks random must have been produced by a random source. But, logically, this does not follow at all. One can suggest (indeed, given our experience of freedom, one must suggest) that the source is a free agent just like us, who determines what from the outside looks like the random effect but from the inside (i.e. in our experience of life) is a sovereign determination. Perhaps the confusion derives from the fact that when one thinks about free will “from the outside” as it were, the all-important “agent” is simply not there.
Anyway, it is clear that even many theists have found that standard argument convincing, and are therefore troubled by the idea of free will. Perhaps the reason that I have not fallen for that argument is that I am a rare subjective idealist (don’t ask), and therefore am used to thinking only “from the inside” as it were. (When discussing with others I mentally translate my views into dualistic language.)
In the case of theists there is perhaps another factor at play. During centuries science appeared to say that physical reality is deterministic, so that all events, including peoples’ choices, could not have been different than how they in fact are. Many knowledgeable theists (who never consider a scientific discovery as some kind threat, but simply as one more insight into the way God has created the universe) accepted determinism as a given. Moreover some notions about God’s omnipotence and omniscience appeared to say that we can’t possibly act in a way that is against God’s plans, and that divine plans are fixed and precise. On these theological grounds then it follows that the world must be deterministic, which is exactly what science was then saying. Thus the dogma of predestination evolved, and some theists found themselves in bed with naturalists trying do describe how freedom of will can be compatible with determinism, and how one is responsible for one’s choices even though one couldn't possibly have chosen differently.
Dianelos
ReplyDeleteI would continue to argue that a sovereign determination that is independent of deterministic factors is random, insomuch as it can not in any way be affected by the circumstances of the decision to be made, as to be thus affected is to be included in the conditions of the probability function that by your model it then resolves into a single decision.
In this respect it does indeed logically follow that this collapser is random, at least to the extent that we define random as being independent of the probability distribution, or disconnected from all the relevant conditions. This disconnection would appear to make this aspect of agency blind to the circumstances of the decision.
Now, you state upon occasion that you do not see the contradiction, and perhaps I should give up at this point because I think I've exhausted my meagre capacity to articulate it. I suppose I will continue to be perplexed, along with a good many philosophers, until somebody can explain to me how an agent entirely unaffected by knowledge of the circumstances of a decision, can nevertheless make a non-random determination. (And to repeat, because any knowledge they have is already included in your probability distribution, this must be the case). What would non-random even mean in this case? This does not appear to be affected by the perspective one takes, it appears just to be a logical conundrum that those attached to logic are forced to pay grudging respect to.
Bernard
Hi Dianelos,
ReplyDeleteYou appear to see quite clearly something that completely eludes me (and Bernard I think). I assume there must be some basic assumption or something that's missing. In fact, determining what I would need to understand you may be as interesting a problem as the core issue itself.
You say this is not a mechanical question. Ok, but surely we can describe the decision process using some logical model? Agent causation seems to require a break in the causal chain, right? Meaning that what happens at some point is not caused (in some sense) by what happened before (or the “context”).
So there is some step that is neither caused nor random. Here, by cause, I understand all the various inputs, influences, deterministic mechanisms, everything that we can identify as having some bearing on the decision – the rationale being of course that, given a replay of the process, the same “causes” would produce the same effect (putting aside random factors that, I think, are not really relevant to free will).
And then what? We go from state A (input) to state B (decision) in some process that has no reference whatsoever to A. Can you see the problem? B is the result of an act of free will but this decision cannot act on anything relevant to it. Because, if it did, the causal chain would not be broken.
Do I get it right? We pass from state A to state B by a process that does not refer to A at all (broken causal chain). How can this be a decision based on A?
(re-posted again)
ReplyDeleteHi Bernard and JP,
This is getting kind of interesting, because either I don’t see a problem everybody else is seeing, or else I see a solution but fail to explain it.
Let me try this track: As it happens I have worked professionally with encryption technology, where the concept of randomness is very important. Randomness is a mathematical term, and contrary to what many people think the concept of “random data” is incoherent, strictly speaking. It’s not that there are no “random data”, but that the idea of “random data” means nothing. So consider the following two sequences of data:
a) 111110000011111000001111100000111110000011111
b) 011101001010100100101000100111000011010110100
Any proposition of the form “this sequence of data is random” or “this sequence of data is not random” is literally meaningless and reveals a misunderstanding about what randomness is. What is well defined and may exist are “random sources”. (A random source produces data that are ultimately not compressible.) Random sources exist in the mathematical realm and may exist in reality. A random source could produce any of the two sequences above, indeed may produce any sequence of data whatsoever. Non-random sources, for example deterministic, can also produce any finite sequence of data, including the two above. So, when discussing data and randomness or determinism the type of meaningful claim one can make is of this form: “Given data D as well as a set of assumptions A the probability that D is produced by a random source is X” or “Given data D as well as a set of assumptions A the probability that D is produced by a deterministic source is Y”. In practical terms when one studies a long sequence of data there is a battery of statistical tests the result of which is either: 1) The data is very probably produced by a deterministic source, 2) There is no evidence that the data is not produced by a random source. For example these tests when applied to the sequences above would say that “(a) is probably produced by a deterministic source”, and that “There is no evidence that (b) is not produced by a random source”. The latter claim can be expressed also by “(b) appears to be produced by a random source”, or, for short but quite confusingly, “(b) is random”. So these are the limits of our knowledge as far as randomness goes.
[continued in the next post]
[continued from the last post]
ReplyDeleteComing back to our context, when JP writes “So there is some step that is neither caused nor random.” what I think is meaning to say is this: When we observe all the data available related to choices, and remove from them those data which according to our understanding about the process of making choices are deterministic, then we are left with a sequence of data that appear to be produced by a random source. But one cannot go form “data D appears to be produced by a random source” to “data D is produced by a random source”. That’s the erroneous logical step in the standard argument against free will.
If you read back you’ll see that I have immediately conceded that from the outside, i.e. as far as observations go, the hypothesis that some element (or “step”) in the decision making process is produced by a random source, is entirely viable. But this does not comport with our experience of free will, and the only problem we are dealing with here is our actual experience of free will. (Observations are dealt with by science, and here the assumption that that source is random is the simplest one, and also entirely adequate in the scientific context). Thus, I say, given our experience of free will the reasonable thing to do is to reject that hypothesis and affirm that there is an element in the decision making process that appears to be produced by a random source but is not. What name should we give to that element in our decision making process? “Agent” is as good a name as any. The agent then, via “agent causality”, is what causes or determines the random looking effect in our decision making process.
[continued in the next post]
[continued from the last post]
ReplyDeleteSome more comments which may help clarify what I mean:
JP writes: “Agent causation seems to require a break in the causal chain, right?”
Right. Agent causation *starts* a new causal chain. That’s why I have been speaking of a “sovereign” and “creative” element in our decision making process. (I think I also wrote of a sovereign and creative final call – which may have been confusing. The final call is sovereign and creative in the sense, and to the degree, that it includes a sovereign and creative element in it.) And that’s why it is fair to say that we, to some degree, are uncaused causes ourselves.
Incidentally, I am not the only one who has noticed the above. Atheist philosopher Tom Clark writes in www.naturalism.org “[…] such causally privileged freedom is a characteristic of God – the uncaused causer, the prime mover, who acts without himself being at the effect of anything. The assumption of free will, so widespread in our culture, in effect sets us up as supernatural *little gods*, and it’s this assumption that a thorough-going naturalism upsets.” Typically enough, he goes one to say that one should not believe this because “there’s no evidence for it”. He is right in pointing out that there is no observational evidence for it, but he misses the fact that the experiential evidence for it is simply overwhelming. Naturalism, it seems, is based on an epistemology in which one keeps one’s eyes tightly shut when it comes to one’s subjective experience of life. On the other I think it’s impressive that he gets right something that many theists fail to see: That free will makes people share a divine “supernatural” property, which I would say is part of the meaning of the famous theistic claim that “we are made in the image of God”. Free will, as Tom Clark rightly observes, makes us into “little gods”, in that we are to a small degree (which accumulatively may become a large degree) unmoved movers, and creators out of nothing. Which conversely may illuminate the theistic claim that “God is not less than a person”.
“We go from state A (input) to state B (decision) in some process that has no reference whatsoever to A.
Exactly right, and that’s why theistic reality is intrinsically and fundamentally creative. How boring it would be if nothing genuinely new could possibly obtain.
Hi, Dianelos-
ReplyDeleteRandomness does seem to be a very foundational point for you, since you also base the possibility of god more generally on the interstices of randomness- that quantum physics invites all manner of interventions and "presences" of god that would not be visible to us due to our interpretation of its action as "random".
If data looks random, then you can spin all you want, but it doesn't support any further hypotheses, especially if it is, when looked at very hard, resolutely random, with well-characterized and stable probability distributions. You are offering a classic god of the gaps argument over a very broad scope- for free will, for biology, and for theistic meddling, or "support" of reality in general.
And you read into this gap the most ornate hypothesis- that some entirely other reality (supernatural) containing an amazing being operates through these random blips on our current world, to guide it to ... to where evolution and the other natural processes would guide it anyhow? It seriously begs belief, to put it simply. Our forebears wouldn't have put up with it for a moment. They were swayed by signs and wonders, not by sophistical re-interpretations of the regular, mundane everything-is-going-to-hell aspects of their world.
What, in the end, is the difference between your theory and naturalism? None, other than your evident conviction that the randomnesses and other natural processes we are subject to come from some beneficent cause and are taking us to some glorious end. All wishful thinking, as far as logic can reach, but essentially harmless if the basic physics is not disputed.
Hi Dianelos,
ReplyDeleteYou agreed with the following: we go from state A (input) to state B (decision) in some process that has no reference whatsoever to A.
Now, as Bernard points out, “no reference whatsoever to A” implies that the free decision, the creative act, cannot be about A at all. That is, if you have to make a decision about some situation (which is part of the initial state A), the free part of the decision cannot be about the situation itself. In effect, then, free will becomes irrelevant to the decision process.
However, perhaps you understand “no reference whatsoever” in a different way.
I think that if we can clear this up we'll have made significant progress.
Hi Burk,
ReplyDeleteYou write: “Randomness does seem to be a very foundational point for you, since you also base the possibility of god more generally on the interstices of randomness- that quantum physics invites all manner of interventions and "presences" of god that would not be visible to us due to our interpretation of its action as "random".”
First a point about nomenclature. The logical negation of determinism is not randomness but indeterminism. Randomness and indeterminism are very different concepts; randomness implies indeterminism, but indeterminism does not imply randomness. If you flip a fair coin a thousand times and count the frequency of heads the result is indeterministic but is not random; indeed it is quasi-deterministic as it will be close to 0.5. The following definitions will do: Given the state of a system and N conceivably possible states of it at a future time, and given that each of the future states will obtain with a particular probability then: Determinism holds if all futures states have probability 0 except one which has probability 1 (and thus will certainly obtain). Indeterminism holds when determinism doesn’t, i.e. when there are at least two possible future states with probabilities greater than zero. Randomness holds when all future states have an equal probability.
In indeterministic systems the concept of randomness enters as a model of *how* it comes to pass that one out of the various probable future states will obtain, or, if you prefer, what causes one rather than the other future state to obtain. Suppose, for simplicity’s sake, that there are only two futures states with probabilities 75% and 25% respectively. Then the idea that nature employs a mechanism which can be visualized as follows: Nature puts in a bag 75 white marbles and 25 black marbles, and *randomly* (i.e. in a way in which each of the 100 marbles has the same probability of being picked) picks one. If it is white then the first future state is realized, otherwise the second one. My whole contention with Bernard is that there is another viable model for visualizing that process, namely one what causes one or the other future state to obtain is not a mechanism but a personal agent who makes a sovereign determination. This latter process is called “agent causality”. Thus, observations which appear to be produced by a process which includes a random source, can also be produced by a process which includes an agent.
[continued in the next post]
[continued from the previous post]
ReplyDeleteNow back to your observation that *indeterminism* is a big point for me. Indeed it is.
First of all because determinism, which one can plainly see is a very bad idea just by thinking about it, has been accepted during centuries both by philosophers and by scientists (including by Einstein in the 20th century) apparently on no better grounds than that scientific models of phenomena at the time were deterministic, and hence that the simplest model for the physical universe was a deterministic one. Einstein until his death believed that there must be something wrong with QM because of its indeterministic nature. But today scientists are trying to absorb deterministic general relativity into QM’s indeterminism, rather than the other way around. More amazing still, when science has now for all practical purposes and for many decades proven that the physical universe is indeterministic, more than 50% of philosophers still believe that it is deterministic. Which only proves the remarkable inertia that exists in the world of ideas.
Secondly, and more importantly, because the physical universe does not only turn to be indeterministic (which is anyway the by far more probable property among all possible universes) but turns out to be indeterministic in a way that makes the physical closure of the universe logically compatible with theism and its claims, not only about our free will, but about God having actually designed the human race to be rational and attracted to goodness, not to mention about God playing an active and important role in human history. I mean how amazing is that? That of all possible naturalistic universes ours should be such that all scientific knowledge (i.e. all knowledge about the physical universe) *cannot possibly* give us evidence for disbelieving in the wildly extravagant theistic story. It looks like God on the one hand did not want to make His/Her presence obvious (which would invalidate the value of faith, i.e. the value of committing oneself to goodness without being forced), but on the other hand did not want to give His/Her rational creatures grounds for actually disbelieving in Him/Her - and designed the universe in a way to fit with both these purposes. A way, incidentally, which all the smart theists of the past did not foresee even as a possibility, until 20th century science actually revealed the special indeterministic nature of the universe, namely one where physical closure and massive supernaturalism are logically compatible. From where I stand this does look like an amazing, I mean amazingly smart, piece of design.
Hi Dianelos
ReplyDeleteMy responses are being dropped at the moment, but no matter. JP makes the point I wished to very succinctly. And in your response to Burk you capture well your own approach, that there is this thing called agent causality that sits as an alternative to either deterministic or random decision making.
Your notion of agent causality, because it has no reference to the conditions of the decision, appears to be deficient. That is, to use Burk's phrase, when we look under the hood, we find it is missing the will aspect of free will.
Your agent causality must act, according to your model, with no reference to beliefs, desires, knowledge, expectations etc. It is a blind agent. Now a blindly acting agent and a randomly acting agent do not seem to me to have any significant differences, which is the point I have been perhaps clumsily trying to make all along. So, along with JP's clarification, you could perhaps explain what the difference is.
Finally, I think you're being a little harsh on deterministically minded philosophers. I tend to agree with them, I suspect random processes will turn out not to play a significant role in the working of the brain. Time will tell, but that there are indeterministic processes within the universe does not imply that all macroprocesses will be driven by these random fluctuations. In fact, our experience of the macro world shows the extent to which such fluctuations wash out in the larger mix. So with regards to brain function this remains an unresolved empirical issue and in the absence of any mechanism by which these quantum effects play a role (I know the likes of Roger Penrose have put controversial hypotheses forward) brain determinism appears at least a reasonable hunch.
Bernard
Hi Burk,
ReplyDeleteA few more peripheral points. You write: “What, in the end, is the difference between your theory and naturalism?”
I am actually impressed that after all the discussions we’ve been having you should ask this question. So let me list a few of the bigger differences:
1. Naturalism has a huge problem with consciousness; theism doesn’t.
2. Naturalism denies free will; theism affirms it.
3. Naturalism denies personal responsibility; theism affirms it.
4. Naturalism denies that some deeds are intrinsically good or evil; theism affirms it.
5. On naturalism the deeply mathematical nature of the universe must be understood as some kind of miracle; on theism it is just what one would expect.
6. Naturalism has even trouble making sense of modern scientific discoveries (such as the fine-tuning of the fundamental constants, the nature of quantum phenomena, the computationally complex behaviour of physical primitives, etc) and therefore finds itself proposing ever more wildly implausible not to say magical worldviews; theism does not have any problems at all.
It seems to me that to be a critical and self-conscious naturalist is to be capable of embracing a deeply unnatural and paradoxical worldview without any good reason at all.
Hi Burk,
ReplyDeleteOne last peripheral bit I wish not to ignore. You write: “All wishful thinking, as far as logic can reach, but essentially harmless if the basic physics is not disputed.”
If theism is true then theistic claims *must* sound like wishful thinking, so it’s only normal that generally they do sound this way. (It’s only some theistic claims which do not sound like wishful thinking that have a problem of coherence.) Further there is the well-known “genetic fallacy”, which many naturalists fall for. They think that if they give a naturalistic account about how theistic beliefs are formed they are making an important point (see Daniel Dennett’s “Religion as a Natural Phenomenon), as if any theist has ever suggested that God has created nature in such a way that believing in His/Her existence would be an unnatural thing. Of all the bad arguments that many atheists uncritically keep in high esteem, the “wishful thinking” argument must be one of the worse.
Hi Burk,
ReplyDeleteOne more and definitely last observation (that post of yours gave me quite some food for thought). You write: “If data looks random, then you can spin all you want, but it doesn't support any further hypotheses, especially if it is, when looked at very hard, resolutely random, with well-characterized and stable probability distributions.”
A generally accepted epistemic principle is that if X *seems* to be the case then one should believe that X *is* the case – unless a good reason (a “defeater”) is there that evidences that X is not the case. So, for example, given that the Earth seems to be flat one should believe that it is flat, unless one encounters a defeater. And in this case there are many defeaters, such as that when one approaches an island from the sea one first sees the peaks of its mountains, then its hills, and lastly its coast. Another example: If one sees what seems to be a red apple on the table one should believe that there is a red apple on the table unless some possible defeater is presented (e.g. reasons to believe that a sophisticated holographic projector is producing that experience, or that one’s brain has been manipulated so that one experiences as red what in fact is green, etc).
Applying that epistemic principle one might argue that our choices, to the degree that they are not determined by the various relevant reasons, seem to be produced by a random source and that therefore one should believe that they are produced by a random source. But, of course, there is a huge defeater for this belief, namely that it implies that one has no personal responsibility for one’s choices. This is an absurd implication of it says that Hitler has no personal responsibility for his decision for the Holocaust, or that when a person helps a stranger in need then that person has no merit for her choice either. In both cases, some random source happened to move some bits of either person’s brain this way or the other.
What’s more, we are not discussing observations about our choices but free will. Thus we should assume that how free will *seems* it also *is*, unless we find or are presented with a defeater. And free will does seem to be determined by us and not by some random source. Bernard argues he has such a defeater, but, in my judgment, he is trying to “look under the hood” of how the agent makes the final call (after the probabilities are set) whereas I say the agent’s process of determining the final call is irreducible and thus that there is no hood to look under. More about this later.
Hi JP,
ReplyDeleteYou write: “That is, if you have to make a decision about some situation (which is part of the initial state A), the free part of the decision cannot be about the situation itself.”
Exactly right. If the free part of the process were caused by the previous state A it wouldn’t be free. And that’s how the application of freedom brings genuinely new stuff into reality, it creates something out of nothing.
“In effect, then, free will becomes irrelevant to the decision process.”
Not at all. The agent, by making a sovereign decision, *determines* (within the available space of freedom allowed by the probability distribution) how state B will turn out. (Please observe that the concepts of “free will” and “agent” are not independent: the existence of free will entails the existence of an agent who applies it.) It is the presence of the agent’s determination between A and B, and the agent’s power to be an uncaused cause, which introduces into B something that has indeed no reference whatsoever to A, and thus is completely new.
That’s the beauty of it. We exist in a world, which despite all the limitations we suffer from, gives us space of true freedom, in which we have the sovereign power to act, to change the world and to change ourselves in ways that transcend the past. (Further by slightly changing ourselves also, the power of freedom is cumulative.) Free will then, when used wisely, has the potential of liberating us from the past, including our own past, and makes us co-creators of ourselves. Which in turn makes us responsible for our state, however it is. (Actually a deeper analysis will reveal that one is more responsible for one's good state than for one's evil state. In other words the personal merit of good people is greater than the personal guilt of evil people. The reason for this is that a good state tends to increase one's freedom and an evil state tends to restrict it. Good people tend to be masters over themselves; evil people tend to be slaves to their past.)
I say, the view I expound above is not only entirely compatible with theism (and indeed with what I hold to be the strongest theodicy, namely Irenaean theodicy), but also fits exactly with how we experience life. I say that at least we adults can’t help realizing that we are to a large degree responsible for our current personal state. Perhaps I should qualify this: I feel to a very large degree responsible for my own personal state, but I suppose there may be horrific external conditions or events (Eric calls them “soul crushing”) which absolve some people from being largely responsible for their own personal state. It’s not that these latter people (who I hope are few) are not responsible for their choices; rather it’s that the space of freedom that life opened for them was such that their choices lost the power of the cumulative effect. To use the sailor analogy, it’s the case of the unlucky sailor who fell into such a storm that his actions were too restricted and the course of his boat was determined far more by external factors than by the sailor’s will.
Eric,
ReplyDeleteI wonder whether the consideration that by becoming a better person one gains freedom, and by becoming a more evil person one looses freedom, and that therefore good people have great merit whereas evil people have little guilt – is not something that can be used against the idea of hell: Namely that whereas good people merit paradise, evil people do not merit hell.
Another related implication: Whereas good people deserve our admiration, evil people do not deserve our contempt, but rather deserve our compassion. It is rarely the case that a person became a good person because life happened to give her all the good motivations, so it’s rarely the case that a good person has little merit. But I think it’s not rarely the case that a person became an evil person because life happened to give her all the bad motivations, so it is not rarely the case that evil an evil person has little guilt.
In short, it seems to me that the paths towards goodness and towards evil are *not* symmetrical, but of a very different nature. And that this has deep implications for theistic eschatology.
Dianelos
ReplyDeleteI still struggle to understand how your version of freedom, this uncaused cause, is in any sense wilful. It appears to me that you are defending, and wrapping in poetry, the beauty of a random twitch. You concede that the final call is independent of the conditions under which the call is made. That strikes me then as an entirely empty capacity. Am I missing something?
Let me put this in concrete terms. I walk into the voting booth, with all my knowledge, moral sense, prejudice etc on tap, and have a probability distribution leading me more heavily towards one candidate than another. Then, my free will kicks in and I chose a vanishingly unlikely option, setting fire to the polling booth in an undirected and pointless protest.
When asked why I did it, I can not explain, for my free will made me do it, and this free will sits outside causes, it is inexplicable. The best I can do is explain why the act I committed was extremely unlikely. Unfortunately, in this particular case, my free will chose the unlikely outcome, for no reason at all, because that's how your free will works, independently of reason.
This apparently gives you moral responsibility. What a strange form of responsibility this is. I say it gives you freedom but no will, and hence no responsibility at all.
You say that free will is irreducible and it makes no sense to want to look under the hood. That's fine, but we can still, by relying upon your definition of free will, look at all the things free will is not, the things that can't possibly be under the hood. In this case a relationship between a choice and the reasons for making the choice. And so your problem remains.
Until you are able to address this, the claim you make to Burk, that theism can explain free will and moral responsibility, is perhaps dishonest. Similarly for the claims that theism explains consciousness, or that the universe is mathematically complex. Our discussions on these topics show at the very least that they are highly contestable. Your certainty in the face of unresolved debates is the very thing that frightens me away from theism, it seems to have the capacity to shut down intellectual curiosity with grand pronouncements.
Agnosticism on the other hand says, let us see mystery for what it is, a chance to explore with an open mind. Isn't that just a much more optimistic, exciting and ultimately humble point of view?
Bernard
Hi Dianelos,
ReplyDeleteYou agree that the free part of the decision cannot be about the situation itself.
So, the creative act of the free agent has nothing do to with the facts of the case. It is not influenced by them, doesn't consider them in any way. It only chooses one outcome, constrained only by the probability distribution.
Then, of what use is this free agent? We could replace it by a random choice and we would not be any worse. There is no difference that I can see.
jp
Hi JP,
ReplyDeleteYou write: “ You [Dianelos] agree that the free part of the decision cannot be about the situation itself. So, the creative act of the free agent has nothing do to with the facts of the case. It is not influenced by them, doesn't consider them in any way. It only chooses one outcome, constrained only by the probability distribution.”
Perhaps you and Bernard project an artificial division on my account of free will. I am not saying that there is an agent who considers all the issues related to a decision and then a second agent who handles the free dimension of it. It’s not like the first agent works up to the point that a probability distribution is fixed, and then passes the buck to a second agent, who is blind and unthinking and who mechanistically makes a random choice. That’s *not* my account at all.
Rather, in my account, the *same* personal agent who considers all factors, experiences their conflictive pulls, thinks about the issues, settles to some probable choices – is also the agent who makes the final call within the space of freedom that is now available. And *that* final call is not reducible or contingent on anything else, but is rather sovereign and creative, determined by the personal agent alone, without using any kind of “mechanism” whether random or deterministic or whatever. It is here where we become co-creators of the world, and thus of ourselves too. This is one more place where we are like God.
To understand my account you must be capable of at least visualizing the possibility that part of our nature is divine and not subject to physical contingency or order. That personal nature is not reducible to mechanisms. And therefore that to understand a personal being is ultimately to understand that being in the personal/poetic terms we use for speaking about our personal acquaintances, and not in terms that describe a mechanism. Actually, isn’t that kind of obvious? Do we understand the people we are acquainted with, and especially the people we know best and thus love, in terms other than poetic?
“Then, of what use is this free agent? We could replace it by a random choice and we would not be any worse. There is no difference that I can see.”
Yes, there is no external observational difference. But there is of course the huge difference we’ve been discussing: That if it is a random source rather than the agent who determines the final “free” call then there is no personal responsibility for one’s actions. I can hardly imagine a difference more momentous than this for our understanding of ourselves and for how we view our lives.
Dianelos-
ReplyDeleteYour definition of indeterminism is fine. I was alluding to the same thing by mentioning probability distributions. Within a given distribution, one's outcome is random, meaning that it can not be predicted by analytical means as far as we know. How you can turn that into "a sovereign determination" is beyond me. I don't even understand the point, but to save appearances for theism. It is like saying that electron distributions around a hydrogen atom are run by "sovereign determinations" rather than random chance, yet are no more predictable or "sensible" for all that, following strictly in their random way.
"Thus, observations which appear to be produced by a process which includes a random source, can also be produced by a process which includes an agent."
Yes, well- wouldn't Occam's razor apply? Note that by my perhaps naive understanding, your agent is typically not imputed to to be running the distrubution all the time, but only in the few times required to intervene for the good of evolution, morality, human history, etc. Thus you would need two mechanisms in place of one. If it were running the distribution of every electron all the time, it still begs belief as an overall theory, there not being any backing point or evidence for such a omni-puppeteer required, other than, again, to save theism as an analytically meaningless abstraction.
"... turns out to be indeterministic in a way that makes the physical closure of the universe logically compatible with theism and its claims, not only about our free will, but about God having actually designed the human race to be rational and attracted to goodness, not to mention about God playing an active and important role in human history"
This simply fails to follow. The fact that, as above, you could way over-interpret what appears random/indeterminate to the critical observer as allowing room for whatever spook you desire to impute, (as long as no one is looking), without logical contradiction ... hardly makes it logically "compatible" with all the rest. This is like saying that 1+1=2 is logically compatible with the evolution of plants. They may occupy the same universe, but their logical connection ends there. To make an argument, you would have to either present a counter-factual theory of how reality would look without the indeterminism into which you pack so much theism, or evidence that the known indeterminism really does take an active hand in our affairs, reported in a more sober fashion than managed in the scriptures.
"I mean how amazing is that?" What is amazing is your capacity to way, way, way over-interpret your remit, and to twist the absence of evidence into evidence for well-worn mythological tales.
cont...
ReplyDelete"... This is an absurd implication of it says that Hitler has no personal responsibility for his decision for the Holocaust ..."
This is the frequent theist lament. But my theory of no-free will has no problem with responsibility. I would say that Hitler was responsible precisely to the extent that he had learning and cognitive capacity to see harm come from his ideas and action, and failed to remedy that harm, as guided by a normal conscience and cultural norms. We don't convict cats for killing mice, and we don't convict the mentally incompetent for crimes they commit- because they are not responsible beings. They have free will, but they don't know the standards/rules of cultural behavior to which they are responsible.
Now if the Nazis had won and changed the world's norms, we would be in the different position of holding Hitler personally responsible for the greatest flowering of cultural rejuvenation the world has ever seen, or something of the sort. And naturalists would be willing to assign responsibility there as well, recognizing through all the vageries of history and culture that the personal component is only a part of the story, but recognizing as well that if we want more of that great leadership, we should cheer it on and thereby reinforce it by social approbation.
Responsibility is an operative social concept, not an ontological one. It is an important part of our instruction mechanism telling people what is good, what is bad, how they should act, etc. Humans are cultivatable beings, able to learn and change their reactions. The fact that all this takes place on a substrate of natural mechanism doesn't alter the flexibility of the algorithm, and thus the responsibility we give it, when it is functional.
Hi Dianelos
ReplyDeleteI still think you're dancing around the central problem.
I'm not imposing any artificial division, as you suggest. The only division is the one you make in your theory. Here's how I understand it, tell me if I'm mistaken:
There is one agent, processing all the pertinent information but finding this mechanistic processing does not of itself yield the decision. Rather we get a set of options, a probability function if you will. This same agent has a transcendent, non-mechanistic process available to it that, taking these probabilities into account, freely makes the decision, so determining the way the decision falls. Because the probability function holds, these decisions must appear random to the outside observer.
Or, in your words:
"And *that* final call is not reducible or contingent on anything else, but is rather sovereign and creative, determined by the personal agent alone, without using any kind of “mechanism” whether random or deterministic or whatever."
So this final call operates in splendid isolation from the factors that set up the probability function. It makes the final call for no reason whatsoever; not on morals, not on tastes, not on biological urges.
Again, a concrete example may help you see my problem. I walk past a small child crying in a mall. It has lost its parent. My probability function disposes me mightily towards stopping and offering assistance. Sure, I am in a hurry and busy and maybe somebody else will stop eventually, and this gives the tiniest tiniest probability that I will walk callously on. But my empathy, my sense of duty, my desire to help, these all push me deterministically towards stopping.
But then my free self, my final call, steps in and for no reason at all opts for the unlikely option, as it must sometimes do if the apparent randomness is to hold. It pushes me past the child. This I think is how your model works. It just chooses one or the other for no reason (it is a non-mechanistic process). Mostly it chooses the thing I want to do, in which case it makes no difference at all, but sometimes it chooses the thing I don't much want to do, just because it does.
This occasional subverting of my determined will strikes me as a a strange sort of freedom. Certainly I can't see how it gifts me any moral responsibility at all. Quite the opposite.
Bernard
Hi Dianelos,
ReplyDeleteThere is still a missing element in your explanation and I will try to rephrase it as well as I can.
In understand that, under free will, given the same inputs, the agent may make a different decision (this is the “there is a possible world in which I decide differently” business). The question is simple: on what basis?
The inputs, circumstances, history, emotions, weather, intellectual capacities, whatever you want – everything is absolutely identical in both situations. If not based on a random element, why would the agent make a different call the second time?
A consequence of this seems to be that the agent itself could not justify its decision. Because any such explanation must make reference to something in the input – and the inputs are absolutely identical.
In this respect, the personal responsibility that you see as collateral to free will does not seem to mean much. How can we square this with a decision that has no possible explanation and cannot be justified in any way? How can we be held responsible for something achieved in such a nebulous manner? Couldn't we just say: “sorry, against my better judgment, it's my free will that made me do it?” Bernard's example of the stranded kid becomes very pertinent.
Hi Burk,
ReplyDeleteYou write: “Yes, well- wouldn't Occam's razor apply?”
It would apply if both hypotheses were otherwise comparable. But the hypothesis that our decisions, to the degree that they are free, are determined by a random source, is not at all comparable with agent causality, because it implies such absurdities as that we are not responsible for our choices. And as we are not responsible for our choices, neither are we responsible for the consequences of our choices, including how they affect our own character. So, for example, if somebody chose not to study but instead to pass her time playing cards, she is not responsible for her uneducated state. We are all spinning in the wind of random effects. I mean how more absurd can the implications of a hypothesis get?
“This is like saying that 1+1=2 is logically compatible with the evolution of plants.”
I don’t see the analogy. 1+1=2 does not appear to contradict the evolution of plants, so it’s not at all surprising that it in fact doesn’t. Conversely, the physical closure of the universe does appear to contradict the theistic account of a God who supernaturally and massively affects the universe, so it’s rather amazing that our universe is such that there is in fact no such contradiction. I mean, even today, some atheist quantum physicists (e.g. Victor Stenger) are so confident that there is such a contradiction that they argue that the fact that we have not observed any violation in the physical closure of the universe *proves* that God is not massively interfering with it. I suppose even most theists believe that and therefore hold that the only way for God to affect the universe is by suspending physical laws and realizing some kind of miracle. It seems then everybody is seeing a contradiction where in fact there is none.
“I would say that Hitler was responsible precisely to the extent that he had learning and cognitive capacity to see harm come from his ideas and action, and failed to remedy that harm, as guided by a normal conscience and cultural norms.”
And how do you suggest could Hitler have remedied that harm? If reality is deterministic, then it was written since the Big Bang that Hitler would kill 6 million Jews. And if Hitler’s choices were ultimately determined by a random source within his brain, then how, would you suggest, could Hitler have stopped that random source from causing the brain events which led to him ordering the Holocaust?
Dianelos
ReplyDeleteI agree with you that the deterministic model of will absolutely causes a problem for our traditional conception of moral responsibility. So, we are left with two choices I suppose: try to establish an alternative model, as you are trying to do, or try to establish a different conception of moral responsibility, as Burk appears to favour. The third option I suspect is to learn to live with the paradox, a sort of, 'although we do not appear to have free will, we must live as if we do' stance.
I find free will fascinating precisely because I can't choose between options two and three, and indeed would love to see option one rescued for me. The reason I pass it over at the moment is simply that even in its most clearly articulated form (and yours is very clear I think) I can't see how moral responsibility gets in. So, the little child lost in the mall becomes the example I would need an answer to (using Hitler as an example always feels a little bad taste to me, not sure tragedy should be co-opted for recreational purposes).
Let me just try to make Burk's (and I think someone like Dennett's) version of moral responsibility fly, purely for my own elucidation. Imagine I live in a deterministic world and my decision to leave the child alone and crying is determined purely by my values, my selfishness, my self-serving expectation that someone else will step into the breech, and my ability to turn down my empathy when it suits, to shut out the self-talk of conscience that can be helpful in times of moral stress.
Now, on the one hand, I am not responsible in the sense that it was always going to fall this way, your 'since the big bang...' scenario. So that form of moral responsibility, that which stems from the opportunity to do other, is missing.
But, another form of responsibility can be rescued I think. I was aware of the consequences when I made the decision. I was aware I was working against my own conception of my better nature. I let myself and the child down and freely admit this. I, the me-machine alone, carried out the calculation and caused the effect. I was unable to predict in advance how I would act, I felt in charge of the decision the whole time, and nobody from the outside could have predicted either. In this sense I made a choice.
I feel shame after the event, and having been pulled up on it, fully accept the disapproval of my peers. This shame and disapproval in turn will inform my future moral decisions. It has become part of my continuing narrative.
The question becomes, how much of a loss is this notion that, were we to replay the tape, we could have done other? One school of thought is that it's an illusory difference. The point is made that the future is always set in this respect, no matter how free the participants. The future is what it is, a mythical creature able to step outside space time would see the way it is going to be and that is that. The idea of two options being open at point A becomes a beguiling metaphor with no real moral content.
I don't know, but I think your claim that Burk's approach makes a mockery of moral responsibility is too easily made.
Bernard
bernard
I am unable to post in this thread; but I trust JP and Bernard who are still participating in this thread may read my posts on their email.
ReplyDeleteHi Bernard,
ReplyDeleteYou write: “But, another form of responsibility can be rescued I think. I was aware of the consequences when I made the decision. I was aware I was working against my own conception of my better nature. I let myself and the child down and freely admit this. I, the me-machine alone, carried out the calculation and caused the effect. I was unable to predict in advance how I would act, I felt in charge of the decision the whole time, and nobody from the outside could have predicted either. In this sense I made a choice.”
That’s like saying that if the handrail breaks causing you to fall from the first floor balcony on a child and injure it, you are personally responsible because while falling you were aware that you would fall on the child.
I personally judge such intents of trying to establish a different conception of perfectly well understood word such as “responsibility” to be intellectually disagreeable. Daniel Dennett is doing this all the time, but I think it amounts to nothing more than playing with words at best, or trying to deceive (perhaps deceive oneself) at worse. If determinism is true, or if our choices are ultimately determined by a random source, then there is no personal responsibility. It’s as simple as that, and anybody who holds either belief should face up to the implications of them. (Your third option of living with a contradiction is surely possibly but not intellectually proper either.)
Dianelos
ReplyDeleteI think there is an importance difference between the broken hand rail fall and the decision to walk past the child. It is is to do with the fact that my values, self talk, interactions etc determine my walk past, they make up the decision and in this sense I own the call, I am responsible for it. Not so with the hand rail.
Let us not avoid the key point though. If this form of personal responsibility is inadequate for you, and the third option appears intellectually improper, well what choice are we left with? How can we squeeze personal responsibility into your first option? How is it that my decision to walk by (given my probability distribution is heavily weighted against this) not simply forced upon me by the requirement that the unmotivated final call must appear random? I'm still waiting for your answer to this. In its absence, the question becomes which of the three options in more intellectually respectable.
Bernard
PS I don't get your postings via email, so if they haven't stayed up on this thread I've not seen them.
Hi Bernard,
ReplyDeleteLet me start by saying that the fact that you are not “buying” my account has been very useful for me, for your counterarguments have helped me become more aware of the challenges my understanding of free will may be exposed to, and thus has helped me think more critically about my own beliefs. Critical thought is a win-win thing, for either one finds one’s beliefs are not warranted and then one has the opportunity to change them, or else one increases their warrant. Further, your, and JP’s, arguments have helped me understand something that up to now was a mystery for me, namely why so many people have trouble with the concept of free will.
You write: “PS I don't get your postings via email, so if they haven't stayed up on this thread I've not seen them.”
That’s a problem because there is a largish post that I did not manage to make visible in Eric’s blog. I will try again and see if I have more luck.
Hi JP,
ReplyDeleteYou write: “In understand that, under free will, given the same inputs, the agent may make a different decision (this is the “there is a possible world in which I decide differently” business).”
Yes, of course.
“The inputs, circumstances, history, emotions, weather, intellectual capacities, whatever you want – everything is absolutely identical in both situations. If not based on a random element, why would the agent make a different call the second time?”
(Just no nitpick: In real life it’s impossible for an agent to find herself in exactly the same circumstances twice, because the first choice has changed her so she is not anymore exactly the same person. But let me ignore this issue, as I believe it is irrelevant to the gist of your question.)
If the agent makes a different choice the second time, it does so because it has so willed. As we have been discussing, to the degree of space for freedom that each state of affairs (the “inputs”) allow, the agent has the power to introduce something entirely new into reality.
Actually, you are giving me a good point here. We know our condition well enough to realize that if we were given exactly the same inputs we might choose differently. Which falsifies determinism.
[continued next]
[continued from above]
ReplyDelete“A consequence of this seems to be that the agent itself *could not justify its decision*. Because any such explanation must make reference to something in the input – and the inputs are absolutely identical.”
That’s an interesting point. The agent can of course justify much of its decision, precisely by pointing out the inputs which often and to a large degree influence it. But I will agree that one can’t give a *full* explanation, for the “final call” itself, the part of the decision which is free, is sovereign and not reducible to anything beyond being determined by one’s will.
So, do you think there is a problem here for my account, and if so why? And let me just quickly point out the account of random source causality suffers from the same issue, for here too there can’t be a full explanation of one’s decision either. Only in a deterministic world is it in principle possible to give a full explanation for one’s decision, but now “one’s decision” becomes a misnomer (it’s not *one’s* decision, but an implication of the state of the universe in the Big Bang).
Perhaps to insist that everything must have a “full explanation” is a pointless dream. Why should it be so? Indeed, even on determinism, full explanations are only in principle, but not in fact possible, because of the existence of chaotic systems.
“Couldn't we just say: “sorry, against my better judgment, it's my free will that made me do it?””
No we couldn’t. Rather it’s on the deterministic view that one can say “deterministic forces made me do it”, and it’s on the ransom source causality that one can say “the random source that made me do it”. On agent causality one can’t say “free will made me do it”, for it’s incoherent. “Free will” expresses the fact that nothing *fully* determines how I choose, for the final call is mine alone. It’s *mine*, not a matter of reasons, or of some process I use, or whatever.
Hi Bernard & Dianelos,
ReplyDeleteThis idea of personal and moral responsibility (in a sense I'm not sure I understand) is coming back all the time and I can't help thinking it's overrated.
Imagine a world populated by robots, endowed with extremely sophisticated programming enabling them to act more or less like humans. They are programmed to have (or simulate) various human-like behaviours: they have preferences for certain things, interests, hobbies, jobs. They live in homes that they are programmed to enjoy, and so on. The whole thing constitutes in fact a simulation of a human society in all its variety. These robots are purely mechanical devices, running a wholly deterministic computer program.
Now, it may happen that some of these robots – through a difference in their programming or their past experiences – occasionally behave in a way that is harmful to other robots. Because they are very sophisticated machines, the other robots will take measures to protect themselves against the rogue ones. They may restrict their movements or try to reform them, for example.
If the situation happens often enough the robot society may take deterrent measures, associating some actions with a “punishment” so that robots would think twice before acting out. It may happen that a guilty robot argues that it was not entirely “responsible” for a rogue actions because, say, it had a defective part and, once repaired, it would again act in a proper manner – and this one could get away without the “punishment”.
So, you see, it seems that the robot society can succeed in implementing perfectly viable social and justice systems based on some notion of “practical responsibility”. They don't need any free will or morality to do that. And, perhaps, we don't either.
Hi JP,
ReplyDeleteYou write: “If the situation happens often enough the robot society may take deterrent measures, associating some actions with a “punishment” so that robots would think twice before acting out.”
What you write here entails that an evil robot can choose among two possible alternative courses of action, and given the social situation will “think twice” and thus more probably pick the proper course. But this contradicts your premise that the robots are deterministic systems.
“So, you see, it seems that the robot society can succeed in implementing perfectly viable social and justice systems based on some notion of “practical responsibility”. They don't need any free will or morality to do that. And, perhaps, we don't either.”
I am not sure how the notion “need” and “need not” applies to a deterministic system. A deterministic system blindly evolves without “needing” anything. On the other hand I agree with you that a society of intelligent robots (perhaps in a simulated deterministic environment) will most probably produce social and justice systems, will produce the notion of “responsibility”, and will even produce ethical philosophy. But I fail to see how all that is relevant to my claim in a deterministic world there is no responsibility, simply because the concept of “responsibility” makes no sense in such a world.
I find that a situation that often arises in discussions is this: A says that X exists (or does not exist) and B counters by pointing out that it "seems" to be otherwise. What seems to be the case, and what is the case, are two entirely different things. It’s easy to agree about what seems to be the case (for we all experience basically the same kind of world). The question is what *is* the case. And I stand by my argument that in any deterministic reality (whether inhabited by intelligent robots or intelligent humans or intelligent gods) there is not responsibility, no matter how it may seem.
Hi Dianelos and JP
ReplyDeleteSorry to have missed that post Dianelos. Like you I find this discussion tremendously helpful, because I don't yet claim to understand the issue of free will very well.
I am still stuck on this example of passing the crying child by, because as yet I can't squeeze any moral responsibility into your system either. Occasionally, in order to follow this rule of appearing random, my final call self must have me do things I am overwhelmingly predisposed not to do (any apparently random system must upon occasion produce outlier results).
Under your system, mostly I will go where my deterministic self would have taken me anyway, and occasionally I must go in the opposite direction, for no reason that relates to the circumstances of the decision.
When you say there is no reason for our choices, it doesn't meet my sense of personal responsibility somehow. If all my final calls are 'just because' calls, with no reference to my past, my loyalties, my beliefs etc, then for me they are entirely empty. It may be that for you, despite these limitations, it's still enough, and I'd just have to accept this, as we've moved past the reach of reason then, but I can't help thinking you might have another trick up your sleeve.
JP, in your robot example, do the robots believe they have the ability to make decisions, or do they feel they are simply running out the programme? In other words, do they have programmed desires, are they motivated to act in ways so as to achieve them, and do they believe success or failure is in some sense up to them?
The issue of moral responsibility for me is an entirely personal one. I seem to have a need to believe I am not simply running out the programme, and that people around me aren't either. This construction of self is absolutely vital to my personal sense of meaning, even though I can't sustain it logically. I don't assume this is true for others, I'm quite happy to accept that some find no such challenge here, but for me it creates a natural tension.
Bernard
Hi JP,
ReplyDeleteThe point with the robots only helps what Dianelos is trying to say. The behavior of the “Normal” members is every bit as determined as the “Abnormal” ones that are causing problems, so any concept of “Justice” is just as foreign as any concept of “Crime”. If the members choose to take measures such as imprisonment or repair in response, they are only acting in accordance with how they themselves were programmed, and they could just as well send malfunctioning members to the incinerator if that was what they were determined to do.
Hi again Bernard,
ReplyDeleteI did manage to get my posts to appear above, but now am having trouble with this one:
You write: “I think there is an importance difference between the broken hand rail fall and the decision to walk past the child. It is is to do with the fact that my values, self talk, interactions etc determine my walk past, they make up the decision and in this sense I own the call, I am responsible for it. Not so with the hand rail.
On determinism the universe is a web of physical interactions which must necessarily happen in a predetermined order. Now take three cases in which a child is harmed: 1) because the rail broke and you fell on the child, 2) because you drove drank and hit a child with your car, and 3) because you enjoy injuring children. In all three cases the same state of affairs holds: The injury of the child was caused by an avalanche of physical events, some of which passed through your brain. In case (1), for example, events in your skull caused your body to lean against the rail.
[continued next]
[continued from above]
ReplyDeleteIt’s absolutely clear that if an event couldn’t possibly not have happened then one is not responsible for the fact that it did happen. Therefore, to search for a creative way to redefine “responsibility” in a way that makes some kind of sense in a deterministic world is a pointless waste of time, an intent to pull off one’s self-delusion.
“How can we squeeze personal responsibility into your first option?”
Well, if the final call is determined by me and me alone, without the possibility of blaming some process, or operation, or reason, or xyz, for it – then it’s clear enough that I am personally and fully responsible for it, don’t you agree? No squeezing is required.
[continued next]
[continued from above]
ReplyDelete“ How is it that my decision to walk by (given my probability distribution is heavily weighted against this) not simply forced upon me by the requirement that the unmotivated final call must appear random?”
That the final call can be interpreted as being produced by a random source, far from being something “forced” upon you, is a necessary condition for opening up a space of freedom in which you and you alone can make the final call. Or consider the converse: If it were possible to interpret the final call as not being produced by a random source, then it would be possible to interpret it as being caused by something pre-existing the final call. But then the alternative interpretation of agent causality, namely that it was you and you alone who caused the final call at the time the final call was made, would not anymore be possible. In short: For agent causality to be a possible hypothesis, it must be the case that the final call can also be interpreted as being caused by a random source rather than by an agent.
HI Dianelos
ReplyDeleteHere is where I think we've got to:
Intuitive definitions of free will seem to place significant weight on two factors. First, the decision is not determined, it could have genuinely been other. Second, the decision is informed, the circumstances of the decision have some bearing on its outcome.
The model I tend towards (determined) violates the first of these. Your model violates the second (or at least the free component of it does). The interesting question then becomes, can either model rescue some meaningful definition of personal responsibility?
I'm not yet sure whether you are arguing that your model doesn't violate this condition, or whether you are arguing this condition is unimportant, and I am happy to return either point should you feel it is controversial.
In the meantime, I will try to develop my thinking with regard to rescuing free will by redefining it. I know you believe this to be a fool's errand, but if your own model also falls short of the comon definition, then maybe this is a task forced upon us both.
Bernard
Hi Bernard,
ReplyDeleteYou write: “[The second factor is that] the decision is informed, the circumstances of the decision have some bearing on its outcome.”
Agreed.
“Your model violates the second [factor] (or at least the free component of it does).”
I don’t think it does. As you appear to be aware, on my model the decision is informed, and the circumstances have some bearing, indeed in most cases a significant bearing. Only the circumstances do not *fully determine* the decision, for there is always the free component. So – where’s the problem for my model?
Incidentally it feels strange to call my views on free will “mine” or a “model”, because I am only trying to describe our experience of free will, and then accept that experience at face value. Perhaps free will is something fundamental in reality and therefore cannot be modeled, for only something non-fundamental and reducible to more fundamental bits can be modeled.
“if your own model also falls short of the comon definition [snip]”
Do you think my description falls short of the common usage of the concept of “free will”? If so, can you please explain where?
Hi Bernard,
ReplyDeleteI don't know what the robots believe. The point of the example was mainly to question the social usefulness of the concept of moral or personal responsibility. I think some version of “practical” responsibility works as well and is all we need for social life. What remains are the religious use of moral responsibility (to make divine rewards and punishments possible, I suppose) and its personal aspect. I agree that it is an entirely private matter.
Another point of the example might be that these robots are, in every meaningful sense, making choices. They are evaluating the evidence, pondering possible options, considering past experiences, perhaps experimenting as a way to test their ideas, and so on. They are limited only by their capacities and knowledge – as we are, even under “free will”.
Now, what some (well, many) might say is that, no, they are not choosing but “merely” executing a program. (As an aside, it is quite interesting to see reductionism adopted by persons of such different views when discussing this topic – while, in other areas, it is seen almost as a sin.)
Saying that these robots don't choose amounts to saying that insects are not alive because they're “merely” chemistry or that mountains don't exist because they are “merely” atoms, and so on. We recognize higher level of organization over the board – except when it comes to these questions. Why is that?
I think there is a major difficulty when conceptualizing determinism. We have this image (at least, it comes naturally to me) that the results of our actions are decided in advance (from the outside) and that, whatever we do, we are inexorably pushed towards that result. But this image is incorrect: nothing is pushing us in any direction. We are the movement itself, not some entity tossed around by deterministic forces. We need to get rid of theses misleading images, I think, before we can hope to understand clearly what's going on. And, well, it is quite difficult.
jp
Dianelos
ReplyDeleteYes, exactly so. Here is our disagreement in sharp focus at last, I think. Under your model/description of free will, the way the probability function is constructed is not free, it is wholly determined by the circumstances of the chooser and the choice, cranking through some sort of algorithm and producing weighted options.
Now, you argue such a function is not decisive, and there is a free component that collapses the function, turning these options into a single choice. This process of collapse is the only place where freedom comes into play, and it is also, as you describe it at least, entirely independent of the circumstances of the decision. So whether we are deciding on which bunch of flowers to buy, or whether to get behind the wheel drunk, makes absolutely no difference to the way the free element of the decision occurs (this affects only the weightings of the options).
If this is not true, then the probability function no longer holds, as outside influences have been permitted. So, to be free is to be blind, at least under this model.
And this is a violation of requirement that free will is informed by the circumstances of the decision. Here only the determined element is informed. The free element is explicitly uninformed.
And so we are in a bind; unless, as JP proposes, we can speak meaningfully of free choices being made within a deterministic model.
Bernard
JP
ReplyDeleteMy instinct is that you are right here, and this reconceptualising will offer the only satisfying way out of the conundrum. Dennett likes to make the point that even under a free model, the future is still what it is and so the free agent still is pulled towards that future in some sense (so should it be possible to observe the future, we would see the choice the free agent 'has' to make).
He goes from here to make the point that this idea of the future being set is, as you say, misleading and unhelpful. I remain unable to find a satisfying metaphor for this limited form of freedom though (You offer 'we are the movement' but that doesn't help me yet, I'll think on it).
You ask why we don't apply the same reasoning to statements like: ants aren't really alive, they're just a bunch of interacting molecules - Well, for the materialistically minded, isn't this because we recognise alive/not alive as a helpful human construction, a tool for categorising? The implication from your question then, is, do we see freedom as also just a category, a manner of speaking about the way the human machine interacts with its environment, or do we need to see it as something more than that?
This then becomes exactly the same argument people have about morality (a construction, or self sustaining concept?) truth, consciousness and beauty. It might be argued that all the debates on this post swirl about this single point.
I find I can give up beauty and truth easily enough, and morality too although the personal narrative begins to creak at this point. But giving up freedom as an actual concept, rather than a higher level description of a category of physical interaction, that might require some key narrative platforms to go for me, at least if I am to value consistency. But I'm unclear on this last point.
Bernard
(repost)
ReplyDeleteHi Bernard,
You write: “Under your model/description of free will, the way the probability function is constructed is not free, it is wholly determined by the circumstances of the chooser and the choice, cranking through some sort of algorithm and producing weighted options.”
Well, yes and no. The probability function is constructed by many factors, most of whom are personal (one’s beliefs, desires, character, etc). At any particular instant these are given, so, like you say, the probability function is determined. But we should not consider free will in relation to a single choice, but in relation to the whole of the human condition. And here we see that the personal factors are not really “given”, but are rather influenced by *past choices*, each of which had it’s own “free element” where the free agent determines the final calls.
This is an important point, so let me restate it: We agree that the current decision depends both on the probability function (itself determined in part by various personal factors) and the agent’s free and sovereign final call within the space of freedom that the probability function allows. But many of the factors that *now* determine the probability function in turn depend on past decisions, which in turn depended in part on the agent’s free and sovereign calls in the past. [continued…]
[…continued] So freedom enters the process of deciding both via the agent’s final call *now* and, indirectly but to an increasingly significant degree, via the agent’s final calls in the *past*. The more time passes, and the more “wise” (in the ethical sense) the agent’s past free final calls were, the more is the actual decision the result of agent causality rather than of extraneous factors.
ReplyDeleteIf you read way back in our discussion, you’ll see that we have already discussed this issue, namely when I argued that my understanding of free explains *why* we don’t hold very young people responsible for their choices. In their case the process of free will did not have the opportunity to realize to a sufficient degree the effect described above by which one is not only responsible for the current free call, but also for the personal factors which determine the probability function and which sets up the space in which the final call can move. In a sense very young people are not sufficiently “personally formed” to be responsible. Further, I have previously argued, my understanding of free has another interesting (and perhaps not generally accepted) implication: That those who did not use their freedom wisely (in short people who moved towards an evil state of personal factors) are less responsible for their choices.
I hope I am making myself clear. Please let me know if not.
Incidentally, I now see that I am perhaps suggesting a model of free will after all. It’s only the “free final call” itself which is irreducible.
Bernard,
ReplyDeleteI’d like to comment to this bit you wrote: “This process of collapse is the only place where freedom comes into play[snip]”
As I hope I have made clear in my previous posts that’s only true if one considers the current decision and nothing else; if one considers a person’s whole life up to now that’s not true. To elaborate: We agree that the probability function is very important; indeed it’s all that is observable from the outside. And, at least in the case of in some sense “mature persons”, the probability function, which to a significant degree is determined by the current personal factors, strongly depends on the past application of free will, which has influenced these personal factors. A wise past application of freedom produces a coherent state of personal factors (i.e. right beliefs, right desires, right character, etc), which to a very large degree is the result of such wise past application of freedom. Conversely, an unwise application of freedom in the past produces a conflictive and chaotic current state of personal factors (i.e. false beliefs, mean desires, evil character, etc) in which the effect of past choices cancel themselves out.
Hi Dianelos
ReplyDeleteExtending our analysis from a single decision to a collection of decisions through time doesn't appear to help much. The probability at point B is still wholly determined by circumstances at point A, regardless of whether or not these circumstances were brought about in part by previous free decisions. If it is true that the free component of the decision at time A is wholly blind to the deterministic circumstances, then this is true of decisions at each point, and this is the crucial distinction.
The notion of the accumulating affect of freedom becomes a red herring, as under an entirely random or entirely deterministic model the same accumulation occurs, past decisions affect current inclinations. So the only way to get non random freedom in is to establish that any individual act of freedom is not blind to circumstance (the quality that leads to many thinkers describing this as necessarily random). I don't yet see how you can do this.
Bernard
Hi Bernard,
ReplyDeletePlease recall what we were discussing. You claimed that the probability function is “not free”, and I pointed out that the probability function is to a large degree the result of the agent’s past free final choices.
Now you write: “If it is true that the free component of the decision at time A is wholly blind to the deterministic circumstances, then this is true of decisions at each point, and this is the crucial distinction.”
In my model it is *not* true that the agent is “wholly blind” to the various factors which determine the probability function. Quite on the contrary.
“The notion of the accumulating affect of freedom becomes a red herring, as under an entirely random or entirely deterministic model the same accumulation occurs, past decisions affect current inclinations.”
In an entirely random or entirely deterministic model there are no free choices at all, so there is no accumulating effect of freedom.
Hi Bernard,
ReplyDeleteIt's true, I think, that live/not alive and so on are categories we use for their usefulness. There is no such thing as Life or The Chair or, even, The Triangle – they are human constructions. Nevertheless, something happens in living things that is not present at lower levels of organization. Take functions like breathing, or walking or structures like a lung or an antenna. These things are genuinely new. There is nothing remotely similar or even indicative of these structures when we look at the underlying chemistry or physics.
Breathing can no doubt be ultimately described in terms of its constituent parts but there is a real sense in which it is not “reducible” to physics (or explained by it) – there is no “breathing” of any sort at the molecular level. Something fundamentally different is happening.
It seems difficult to recognize the same when thinking about “decision making” or “free will” or “consciousness”. Logically the latter appear to be other examples of emergent properties, totally invisible at lower levels. Saying that we cannot have free will because there is none at the physical level amounts to saying digestion is an illusion because it doesn't exist at the atomic level.
[cont]
[cont]
ReplyDeleteAnother stumbling block in thinking about free will, at least in my experience, is this image of “us” as some independent entity opposed to a “program” that we are running. I don't think this is accurate and perhaps this is a left-over from dualist world views. Instead, perhaps we should say that our existence, the way we act and evolve in our environment, happens to follow rules (we are “rules in movement” to use my image). But these rules exist at a level that does not interfere with (or determine in any significant sense) higher level functions. At the electrical level, computers obey extremely strict rules but that does not prevent extremely sophisticated software to be written – it may even be necessary to enable this.
And, here's a thought, a paradox perhaps. Maybe the fact that our low-level structure is in some sense algorithmic is a necessary prerequisite for freedom because it provides the reliability needed for higher level functions. I have to think more about this – the idea just came to me. But wouldn't that be interesting? Instead of being a show-stopper, low-level predictability becoming the key to freedom.
Dianelos
ReplyDeleteLet's stay on this point, it's crucial I think.
I do not claim the agent is wholly blind, I claim that component of the decision that is free is wholly blind. If this is true, then freedom and randomness do, I think, become synonymous, and the only cumulative affect of past decisions is the blind and erratic functioning of your free agent.
So, I need to be convinced that at a given point of time, facing a given decision, the final call portion of your decision making function can be in some manner aware of any of the factors underlying that decision. And I don't see how we can make that work, because for the probability function to collapse, the collapser has to have this random, or at least blind, quality. Otherwise we have misdefined the notion of a probability distribution I think.
I just can't see how you get around this issue. Once it's resolved we can then look to see if the passage through time of a decision making individual brings new implications to your model (I don't think it will). But let's get the model straight first.
Bernard
JP
ReplyDeleteThis is very interesting. Thank you.
I've never been able to make the idea of emergent qualities work, because it seems to me the only thing that ever emerges is a human centred categorisation.
So, to use your example of breathing, I would have thought breathing is exactly reducible to molecular processes, and the only difference is that in order to focus our attention on some of the macro qualities, we think in terms of respiration rather than in terms of atomic level reactions. So all that ever emerges I suppose is an appearance, a way of thinking.
Applying that to the mind, what emerges is only the appearance of consciousness, the appearance of freedom etc. These become helpful categories for contemplating, anticipating, narrating etc.
At this point the argument about whether we are really free becomes pointless maybe, because freedom isn't that sort of thing. To feel free, to appear to oneself free, is to be free, simply because that's all freedom is, a category of appearance.
If so, your argument may hold, for the appearance of freedom may indeed be reliant upon the stability and predictability of the underlying algorithms, just as the appearance of life is dependent upon the stability of the underlying chemistry.
An interesting thought.
Bernard
Hi Bernard,
ReplyDeleteYou write: “I do not claim the agent is wholly blind, I claim that component of the decision that is free is wholly blind.”
Please forget “blind”. There is nothing “blind” in my description of free will. If you think you see something “blind” in my description, then you are being led astray by other peoples’ description of free will.
Perhaps we have a problem with nomenclature, so let me describe my account again:
There is process we often experience in life, and with which we make a free decision. That’s a given. The “we” here, is also called the “agent”, a notion which allows us to talk about the process of decision making in a general fashion.
Now the agent has beliefs, desires, character, etc, which we call “personal factors”. The personal factors are not external but are an intrinsic part of what the agent is. There are also “external factors”, say, the external state of affairs in which one must make a decision (e.g. if I see a tiger about to attack me that becomes a very influential factor in my next decisions). All such factors delimit a set of viable choices, and their push and pull on the agent’s experience of the situation defines a “probability function”, i.e. with what probability the agent will pick each one of the viable choices. Indeed, when the time comes for the agent (the very same agent I am talking about all the time) to determine the “final call” (i.e. which choice will be picked), the agent, while acutely aware (and not at all “blind”) of the various factors and their pull and push and hence of the probability distribution, creatively and irreducibly (i.e. not using any other *component*, process, operation, reason, or whatever – thus in a state that is “sovereign” and “free”) does determine the final call. Observe that at this point there is not any *component* that may potentially be “blind”. The final call itself is the point where the agent creatively introduces something genuinely new into reality, something that cannot be fully explained by or reduced to the previous state of reality. And, as the final call is part of the whole process by which a decision is made, I call the whole process “free” also.
Finally, in the typical case of a normal adult, the agent is strongly “responsible” for its decision, for the agent is not only fully responsible for its current sovereign final call, but also, to normally a significant degree, for its current personal factors, which set up the current probability distribution.
I see that process for making decisions very clearly. If you see any problem with it, please point it out *without* changing or adding elements or properties to it.
“So, I need to be convinced that at a given point of time, facing a given decision, the final call portion of your decision making function can be in some manner aware of any of the factors underlying that decision”
My mastery of English is not perfect, but this kind of language strikes me as unnatural and obscure. The “final call portion of a decision making function” is *not* the kind of thing that may be aware of something of other. Only the agent may be aware of something or other.
Hi Dianelos
ReplyDeleteThanks for your patience. I sense this conversation is frustrating you somehow, but I hope you are nevertheless getting something out of it. I certainly am.
Your restating of the process of decision making seems to be in line with my understanding and so it is over to me to explain what I mean by this blindness I suppose.
You are reluctant to separate out the setting up of the probability function and the making of the final call, whereas for me the probability function only makes sense if there is an independent process by which it can collapse into a given decision. If there is a way of maintaining the parameters of this function whilst collapsing it using a non-independent process, then this is the vital thing I've missed and your explaining this will help me out of my confusion.
If not, then whilst there is one agent, with an overall sense of the various factors in play, the process by which they make the final call is blind to the circumstances in an important sense. Specifically the circumstances in question can have no bearing upon the way in which the function collapses (beyond of course their influence on the parameters of the function in the first place).
So the agent can be perfectly aware of all the factors leading to the setting of the probabilities, but in deciding how these should then collapse, they must ignore them. If they do not, then the very description of the probability function would be violated I think. So this blindness comes about through your very description of the process, that's where I got the idea from.
In practical terms, whatever process this sovereign power uses to make a final call, be it whether to get a haircut or commit a murder, the process is entirely unaffected by the nature of the decision being confronted. This is what I meant by blindness, and I apologise if that's a misleading metaphor, I can't think of a better one.
With this clarification in mind, you can perhaps explain how either this blindness does not in fact occur, or how it is unimportant for your conception of freedom.
Bernard
Hi Bernard,
ReplyDeleteLike you I am trying to get something out of our conversation, and I feel you are projecting your own assumptions on my understanding of free will. Of course you may hold any assumptions you like, but if my understanding does not fit your assumptions this does not imply that there is something intrinsically wrong with my understanding. What interests me is to find out whether you can see something intrinsically wrong with my understanding.
My general point is that if one seriously considers one’s experience of free will without making any assumptions then 1) free will makes perfect sense, and 2) there is not contradiction with scientific knowledge. I do see that there is a contradiction with any mechanistic understanding of reality, but I do not myself assume that reality is at bottom mechanistic. On the contrary, the fact that my understanding of free will contradicts any mechanistic view of reality is for me powerful evidence that all mechanistic views of reality are false.
You write: “You are reluctant to separate out the setting up of the probability function and the making of the final call,[snip]”
It’s not a matter of reluctance. By “probability distribution” I understand the probability of how the agent will determine the final call. Thus, “probability distribution” *refers* to the “final call”, and one can’t somehow separate the two concepts. If you understand “probability function” as something that can be separated from the final call, then that’s not my meaning. Nonetheless I am kind of curious: What exactly do you mean by “probability function”?
“[cont] whereas for me the probability function only makes sense if there is an independent process by which it can collapse into a given decision.”
Again, in my understanding of free will there is no such “independent process”. Can you give any reasons why you think that such an independent process should be there?
As far as I can see, unless one assumes that everything in reality is reduced or is based on mechanistic processes, there in no reason whatsoever to think that there should be an “independent process by which the probability function collapses into a given decision”. Neither is there in my understanding of free will something that independently exists called “probability function” which “collapses into a given decision”. From where I stand it really looks like you are projecting your own assumptions into the process I describe, and I simply reject your assumptions because I don’t share them. My claim is not that one can make sense of free will while holding XYZ assumptions; my claim is that one can make sense of free will, and based on that sense discover whether XYZ assumptions are viable.
[continued next]
[continued from above]
ReplyDelete“If there is a way of maintaining the parameters of this function whilst collapsing it using a non-independent process, then this is the vital thing I've missed and your explaining this will help me out of my confusion.”
Again, there is no process, independent or not, in what I describe. There is no such thing as an independently existing probability function which in some way “collapses”. You keep seeing things that are not there in my description of how we experience making free decisions.
Perhaps the problem resides in that you try to use the same language to describe the experience of free will, that a physicist may use to describe the physical operation of a brain. For example a physicist may say that the brain is a quantum system, and that the final call is the collapse (according to the Copenhagen interpretation of QM) of the brain’s wave-function into an observable physical state, which is precisely the decision the brain makes (and say picks the green ballot and puts it into the voting box). If that’s the problem, then please consider this question: Why should the language that describes one’s understanding of free will mimic the language physicists use to interpret quantum mechanical phenomena? The only requirement is that there should not be any contradictions between the two narratives (the experiential and the physical). And if I say “here the agent freely and irreducibly determines the final call” and the physicist says “here the brain’s wave-function collapses” there is no contradiction. Do you agree there is no contradiction? That’s a key point. (Incidentally, in the Copenhagen interpretation of QM there is no “process” used in order to achieve the wave-function collapse either. The collapse is supposed to be irreducible and refer to a fundamental property of reality.)
“So the agent can be perfectly aware of all the factors leading to the setting of the probabilities, but in deciding how these should then collapse, they must ignore them.”
That’s so far away from the description I gave, I don’t know what to make of it. Again, I am not here proposing a new and innovative understanding of free will. I am just *describing* my (and our) experience of free will. Surely when you make a final call you are not “ignoring” the various factors over which you have been agonizing, do you?
“If they do not, then the very description of the probability function would be violated I think.”
What the probability function actually refers to is the probabilities of how the agent will choose. And these probabilities have the values they have *precisely* because when the time comes for the agent to make the final call the agent *does* consider them.
Actually any “blindness” in the final call would produce results that contradict reality. Take the example you have previously suggested: If, when making the final call, you ignored the factors that played a role in the decision process, then, given that there are only two choices, you would with 50% probability make the final call to walk past the child. But that’s simply not how things are.
“In practical terms, whatever process this sovereign power uses to make a final call,[snip]”
What you write here is self-contradictory, for if the agent were to use some process for determining the final call then the agent would not be truly sovereign. Never mind that I have never claimed that there is such a process used by the agent to make the final call. On the contrary I have several times clarified that it’s the agent and only the agent who irreducibly makes the final call, without using any process, operation, component, or whatever.
“With this clarification in mind, you can perhaps explain how either this blindness does not in fact occur, or how it is unimportant for your conception of freedom.”
I’d be glad to try to explain anything you like about the process I describe – except of course what is irreducible in it.
I think what is throwing him off is your talk about the "Probability" distribution. Saying something had a certainty probability of happening, at least in everyday conversation, implies it happened as a result of randomness. I also believe in "Free Will", but I always considered it a "Possibility" distribution myself.
ReplyDeleteHi Thomas,
ReplyDeleteYou write: “Saying something had a certainty probability of happening, at least in everyday conversation, implies it happened as a result of randomness.”
I don’t think that’s true. Many people still believe that the world is deterministic and hence that no randomness exist, but they still speak of probabilities.
I’d say that in everyday conversation when we speak of probabilities we refer to cases where it *seems* that some randomness or perhaps some unpredictability is present.
In the context of our discussion here “probability distribution” means that given an agent’s beliefs, desires, character, etc it is more probable that the agent will pick some choices than others. Which is undoubtedly the case.
Hi Dianelos
ReplyDeleteThe thing is, I'm not adding any assumptions to your model. These assumptions exist within your own description. Here's something you wrote for example:
"All such factors delimit a set of viable choices, and their push and pull on the agent’s experience of the situation defines a “probability function”, i.e. with what probability the agent will pick each one of the viable choices."
This sounds like a description of probability as the term is used in mathematics, and with you see as my mechanistic approach flows from this. I am quite happy to consider a non mechanistic, non-process approach but I think the idea that the viable choices have a probability function and from this function a single choice emerges, precludes this. Now perhaps you mean something quite different by the term probability. This would be an excellent time to explain it.
Bernard
Hi Bernard,
ReplyDeleteI have also had my reservations when reading about these higher levels of organizations (or emergent properties). Sometimes, it seemed the writer simply wanted to deny reductionism and replace it with something too immaterial (or mystical?) for my tastes.
I now think both perspectives are valid. It is certainly true, in a sense, that breathing is exactly reducible to molecular processes. But what does this mean, exactly? What we're doing is expressing or describing a structure or function existing at some level in terms of elements of a lower level. This process is clearly recursive: to identify only a few intermediate steps, breathing is reducible to cellular processes and these in turn can be reduced to complex elements within cells, then down to the molecular level, and so on. As far as I know, it is not even clear that the recursion eventually ends. Who knows? Maybe there is only a non ending succession of more “elementary” structures. If so, everything can be reduced to... I don't know, the Nothingness at the bottom? There is a funny kind of beauty in this image.
In any case, while it is true in a sense that we can reduce complex structures to simpler ones, it is also true that they display properties, structures and functions invisible at the lower levels. They can be observed, manipulated, studied. We can do experiments on them, work out some theory. They are no less (and no more?) real than any molecular structure.
What you say about these (they are appearance, a way of thinking) can be said about anything we know, doesn't it? It seems more a characterization of thinking and knowledge in general than of emergent properties. What I prefer to say is that we know by building a model of reality in our mind. Objects in the model are not the real thing, they are symbols, abstractions. But, for the model to be of any use, these abstractions must correspond to something out there. And the “something”, while different form our idea of it, is very real.
Hi Dianelos
ReplyDeleteI may be able to be more clear:
You state that there exists a set of options with attached probabilities.
A decision then occurs, one of the options is chosen. At this point its probability becomes one. This all I mean by the collapse of the probability function, it's your concept, and I'm just trying to understand it.
Now, when we assign probabilities to future outcomes, we are either attempting to quantify unknown but real factors in a deterministic situation (as in when we roll a die) or we are claiming that between the function and its collapse there sits a truly random element, random in the sense that it is independent of the factors that define the probability function.
I think you are claiming the second, but object to the use of random. You say this uncaused cause is not random, so you are introducing some new element that distinguishes it from randomness. I am uncertain what this element is.
When pushed, you appear to say the element can not be explained because it is irreducible. At this point, I wonder how you know it is not random then. You have at least been able to reduce it to the point that it is distinct from randomness.
This distinction I am further claiming, can not be made with reference to the conditions under which the decision is made, as this violates the notion of probabilities being truly open prior to the choice occurring.
So then, what is this defining factor that makes this non-random? This is the part I am having difficulty following.
Bernard
Hi Dianelos,
ReplyDeleteHere's what for me is one difficulty.
In terms of possible worlds, the question is this: given exactly the same circumstances (inputs), down to the most insignificant elementary particle, there are possible worlds W1 and W2 where the free agent will act differently and the difference is not the result of a random choice.
Let's say I am this agent and is asked why I choose differently. There is nothing in the facts of the case that I can use (they are identical). There is nothing in my values (or preferences?) that I can invoke (they also are identical). In fact, there is nothing whatsoever I can say to explain or justify the difference between the two decisions. Nothing at all.
This part of the decision is under no conscious or rational control at all (because then something could be said to justify it). We may have the best reasons in the world to choose option A instead of B but this agent may come at the last second and override us for no reason at all.
Building personal responsibility on this notion amounts to saying the following: I have personal responsibility because there is a part of me, completely outside of conscious or rational control, that intervenes in my decision process in a mysterious, undefinable, unpredictable manner. Really?
JP
ReplyDeleteInteresting. I need to do more thinking about this idea of emergence clearly. It seems particularly relevant when we speak of things like free will and consciousness. Knowing what we mean when we say these emerge at a particular level of complexity level seems important, and sometimes we mean rather less than we assume.
You're right I suspect to insist that just because a quality emerges at a higher level of analysis, it can't be dismissed as simply a matter of appearance, for appearances are all we can deal with in any situation. So perhaps free will emerges as a result of a certain type of complexity in the same sense that locomotion or photosynthesis emerge.
If this analogy holds then we ought to think of free will in terms of its capacity rather than some more ethereal quality I suppose. This is in line with Dennett's idea of looking at how free will evolves. Free will might turn out to be then a capacity to evade coupled with a capacity with self awareness. That we carry a model of the future, and have an ability to knowingly adjust our behaviour in accordance with that model, becomes the whole game.
Dianelos would say this doesn't rescue moral responsibility. But as you've suggested, it's not clear any model of free will can do that.
I still feel I'm missing something with emergence. I'll think on it more.
Bernard
Hi Bernard,
ReplyDeleteThis has been my longest discussion yet on the web. I have learned a lot though, and have learned the way I like, namely by discovery. Anyway, spurred by our discussion I have ordered the recent “Four Views on Free Will”, written by four academic philosophers. I’ve just received it. The first view is about libertarianism (which I hope is my view of agent causality), the second is about compatibilism (which must be about making free will compatible with determinism), but the third and fourth are “hard incompatibilism” and “revisionism” which I have no idea about. I think I will read this book while our discussion is still warm, and if I find anything of interest I will come back and post it here.
Your recent posts do clarify the issues a lot. We have indeed been discussing three models of decision making, each of which really centers in what we have called “the final call”, i.e. the moment between T and T+1 where the actual decision is made.
The first model is the mechanistic/deterministic model, according to which at T the agent has a sense of a “probability distribution” based on its “best guess for a deterministic process”, akin to how our best guess about the probability of the one centillionth digit of pi being 5 is 0.1, even though that probability is in fact either 0 or 1. At T, the conscious agent unconsciously uses a deterministic process which determines the final call, while the conscious agent remains with the illusion that it determined the final call. This model contradicts both free will (as there are not really more than one *possible* choices), and our sense of personal responsibility.
The second model is the mechanistic/random source model, according to which at T the agent has a proper sense of “probability distribution”, namely one which really describes the probabilities of the various choices among which the final call will fall. At T, the conscious agent unconsciously uses a random source process which blindly and unthinkingly picks the final call within the space of freedom permitted by the probability distribution, while the conscious agent remains with the illusion that it determined the final call. This model is compatible with all observations related to decision making, but contradicts our sense of personal responsibility.
The third model is the non-mechanistic/agent causality model, according to which at T the agent has a proper sense of “probability distribution”, namely one which really describes the probabilities of the various choices among which the final call will fall. At T, the conscious agent consciously determines the final call within the space of freedom permitted by the probability distribution without using any process/operation/component at all. Here the agent’s sense that it determined the final call is veridical. This model not only fits our sense of personal responsibility, but also explains widely accepted properties of personal responsibility, such as that children and mentally handicapped people are not personally responsible for their choices, as well as less generally accepted properties, such as that evil people are less personally responsible for their choices. This last view though comports well with universalism, as well as Christ’s “forgive them Father for they don’t know what they’re doing”.
[continued next]
[continued from above]
ReplyDelete“It's equally a little unfair to say you're simply describing how decisions feel. They don't feel like this to me at all. Rather they fell like a deterministic, algorithmic process with a conscious and subconscious component and narrative feedback loops.”
Well, that’s not at all how I would describe my experience of making a decision. Also, if there is a subconscious component then we are not aware that it is there, so it can’t belong to how something feels like.
“Now, when we assign probabilities to future outcomes, we are either attempting to quantify unknown but real factors in a deterministic situation (as in when we roll a die) or we are claiming that between the function and its collapse there sits a truly random element, random in the sense that it is independent of the factors that define the probability function.”
It’s not “we” who assign probabilities; rather the various factors, personal and external, make some choices more probable than others. Also there is nothing between the “function” and its “collapse”. The probability function refers to how the collapse will probably go. But perhaps you mean something else, namely this: When one starts to ponder a decision there are already the various factors that determine a initial probability distribution, but while one thinks more about these issues the probability function often shifts, because some of the personal factors are contingent on one’s thinking about the decision. If so, I quite agree. What’s relevant to the issue of freedom though is the values of the probability function at time T when the agent determines the final call.
“You say this uncaused cause is not random, so you are introducing some new element that distinguishes it from randomness.”
Right.
“When pushed, you appear to say the element can not be explained because it is irreducible.”
No pushing is necessary. The irreducibility of the free conscious agent’s determination of the final call is the foundation of agent causality. Were I to hide that irreducibility I would fail to describe agent causality.
“At this point, I wonder how you know it is not random then.”
I don’t know that at all, and I hope I have never given a different impression. I don’t know that agent causality (and its foundational property of the irreducibility of the final call) is what actually happens. What I am claiming and am willing to defend is: 1) That agent causality is a coherent hypothesis and thus logically possible, 2) That agent causality does not contradict any piece of scientific knowledge, 3) That agent causality fits remarkably well with our sense of free will and of personal responsibility, and thus works vastly better than any other model for our decision making process I know of. 4) That given the above the most reasonable belief for me is that agent causality is true.
“This distinction I am further claiming, can not be made with reference to the conditions under which the decision is made, as this violates the notion of probabilities being truly open prior to the choice occurring.”
I am not sure I understand what you mean here, but I’d like point out that there is no such thing as probabilities existing prior to the event these probabilities refer to. Probabilities always refer to some event and are meaningless if conceptually separated from that event.
Thanks Dianelos
ReplyDeleteWe do have a clear difference here. The difficulty I encounter is that the final call must be independent of the probability distribution. If this is not true then the probability distribution does not hold, it needs to be modified to include whatever type of interdependence exists.
So, for example, if I have say a twenty five percent chance of choosing A over B in some decision, and will use a coin flip to determine it, then this probability only holds if my coin flip (say I'm after two heads) is an independent process. If the way the coin tumbles is however affected by the very conditions that determined A's probability, then we can no longer say there is a twenty five percent chance of A being chosen.
Now, rather than continue to get bogged down in the details of probability and final calls, because I think we are both struggling to get our definitions clear for the other, JP has expressed far better than I have the fundamental implication of this independence.
"Building personal responsibility on this notion amounts to saying the following: I have personal responsibility because there is a part of me, completely outside of conscious or rational control, that intervenes in my decision process in a mysterious, undefinable, unpredictable manner. Really?"
This in essence is the trouble I am also having. How do you rescue personal responsibility given these conditions?
Bernard
Hi Bernard,
ReplyDeleteYou write: “The difficulty I encounter is that the final call must be independent of the probability distribution.”
How can the final call be independent of the probability distribution, when the probability distribution says something about, or is property of, the final call? Observe that in all cases the agent is not completely free to determine the final call; rather the agent can only determine the final call within the space of freedom the probability distribution allows. So, for example, if the probability distribution is 99.99% that I will choose to pay for the book I picked at the bookstore and 0.01% that I will choose to shoplift it, then the space of freedom available to me is very small indeed and it is practically speaking determined that my final call will be to pay for it.
So it’s not the case that the final call must be independent of the probability distribution, and quite on the contrary the final call cannot violate the probability distribution. So, for example, even though I am free and there is no power capable of stopping me from choosing to shoplift if I so will, given the probability distribution (determined by various factors including personal factors) it’s just very unlikely that I will in fact choose to do so.
“So, for example, if I have say a twenty five percent chance of choosing A over B in some decision, and will use a coin flip to determine it, then this probability only holds if my coin flip (say I'm after two heads) is an independent process. If the way the coin tumbles is however affected by the very conditions that determined A's probability, then we can no longer say there is a twenty five percent chance of A being chosen.”
Perhaps you point here is that in the random source model the final call is independent of the probability distribution, and that the same should hold for the agent causality model. If that’s your point then please observe that your example nicely demonstrates that in the random source model the final call is *not* independent of the probability distribution either. After all, as you say, you are flipping the coin two times and if you get two head then the final call will be A instead of B. But you are doing all that *precisely* in order to respect the 25% probability distribution of A. The random source itself is blind and dumb, but the way it is used in order to produce the final call is determined by the probability distribution.
Finally you quote JP who writes: “Building personal responsibility on this notion amounts to saying the following: I have personal responsibility because there is a part of me, completely outside of conscious or rational control, that intervenes in my decision process in a mysterious, undefinable, unpredictable manner. Really?”
In agent causality the decision is consciously determined by the agent, so I am not sure where JP finds in my model “a part of me completely outside of conscious […] control” intervening. As for the rational control it’s intrinsic in the probability distribution, which is itself determined by various factors including one’s beliefs, desires, etc – which in turn are the result of one’s past free choices. Thus in the agent causality model one does use rational control, and further one is responsible of its use.
Hi Dianelos,
ReplyDeleteMy point is this. Consider again, as Bernard does, the case of two possible worlds in which the free agent makes different final calls. Suppose hypothetically that we can ask the agent why the decisions are different.
The agent cannot offer any rational argument because any such must be expressed in terms of the facts of the case – which are identical. Furthermore, were the agent conscious of any element justifying the difference, it could presumably express it. But, again, this is impossible because any such expression must refer to the facts of the case – identical. You also have made clear the the agent is not simply throwing dice.
Thus, the all-important free part of the agent's decision, the one part that can create different outcomes in different worlds, the ultimate source of moral responsibility, appears to be neither conscious nor rational.
Of course everything the agent does that is identical in both worlds can be as rational as you wish. But the crux of the matter is the difference between worlds.
jp
Hi Bernard,
ReplyDeleteYou write: “So yes, absolutely, the way the coin toss is applied is entirely due to the probability distribution. ”
Right. And in the random source model *how* the toss coin is applied is the key means by which the final call is produced.
“The point, as I understand it, is that the likelihood of a coin landing head or tails must not be affected by the conditions underpinning the distribution function. If it is, then the 25% or whatever no longer holds.”
Sure. The random source in all possible contexts is by definition blind and dumb, and is not affected by anything at all.
“So, whatever means is used to generate a final call that respects the parameters of the probability distribution, this means can not itself refer to those conditions.”
Here you are contradicting what you wrote in the first quote above. You yourself have described the need of tossing the coin twice and seeing if it produces two heads as the means for generating the final call. So the means do refer to the parameters of the probability distribution, in fact in the random source model they are fully determined by it.
The fact that in the random source model the random source itself is entirely blind and dumb, does not imply that the means that produce the final call itself is also blind and dumb. In that model the mechanism which produces the final call is highly dependent on the probability distribution and thus on the various factors that determine it, such as beliefs, desires, character, etc. Actually to speak of “probability distribution” is a short way for denoting exactly how the final call depends on these various factors. Perhaps part of the problem is that you are reifying the probability distribution and imagine that it has an existence all of its own. But the probability distribution is nothing but a property of the means that produce an indeterministic event. That’s the case in flipping two coins, or in blindly picking a marble out of a bag, or in the collapse of a quantum mechanical wave-function, or in the final call of the random source model, or in the final call of the agent causality model, and in absolutely all cases of an indeterministic event.
Coming back to the agent causality model, it occurs to me that perhaps what troubles you is this thought: If a conscious agent is aware of (and thus “refers” to) the various factors behind the probability distribution, then when the agent determines the final call the agent is not fully subject to the probability distribution and may thus violate it. If that’s what troubles you then consider that that’s true only as long as the agent *ponders* the issue, i.e. before the time T where the agent determines the final call. For before that time the agent’s conscious deliberation do affect the various factors, and thus the probability distribution, and thus the means which will produce the final call. This, by the way, is true in both the random source and in the agent causality model. The only difference is that in the former model the various-factors-referring means are mechanical (and includes in part a blind and dumb random source), whereas in the latter model the various-factors-referring means are personal.
Hi Bernard,
ReplyDeleteYou write: “So, we could envisage two identical universes, but in one you steal and in one you don't. This ability to sometimes steal is where you manage to inject freedom into the equation, but note the conditions are identical in the two universes, and so don't we conclude nothing motivated the theft in this case. You just happened to be in the Dianelos steals universe.”
Not at all. Rather, the universe happened to be the one in which Dianelos decided to steal.
In the context of free will it’s us who drive reality, not reality which drives us.
Hi JP,
ReplyDeleteYou write: “ Consider again, as Bernard does, the case of two possible worlds in which the free agent makes different final calls. Suppose hypothetically that we can ask the agent why the decisions are different. The agent cannot offer any rational argument because any such must be expressed in terms of the facts of the case – which are identical.”
The facts in the two possible worlds in which the question is made are not identical, because the agent determined a different final call in each one of them. But I believe you mean the facts of the case prior to the decision. So you ask me to describe how the agents in the two different worlds would rationally explain their decisions using only such prior facts.
Strictly speaking they would both first explain at length the various identical factors that inclined them among the various alternative choices. Then one would end with a terse “and in the end I decided for X” and the other with a terse “and in the end I decided for Y”.
Now perhaps your argument is that neither explanation is fully rational according to your standards, because if the first long part of the explanation is identical then so must its conclusion be. If so, I’d like to argue that that’s an unreasonable standard of rationality. After all the purpose of the initial long explanation was to reveal why both X and Y were probable decisions, so it is to be expected that sometimes the account would end with “I decided for X” and sometimes with “I decided for Y”. If all possible worlds the agent would nevertheless always pick, say, X, it would render that first long explanation nonsensical and thus clearly irrational. (Further observe that if there is a problem with the agent causality model here, the same problem exists with the random source problem.)
Actually, I’d like to turn the table and argue that if the reality in which we live is deterministic then the fact that we do give long accounts explaining why both X and Y were probable decisions when they in fact were not, renders us all grossly irrational, in the sense of justifying our decisions on long accounts which are grossly false.
One last psychological observation. Above I describe the strict answer, but that’s not really how we in practice answer such questions. A decision we have made affects many of the various factors (personal and external) that have previously affected the decision. When we answer the question we shall normally take these new values into consideration, or, in other words, we wont’ end our explanation with a terse “in the end I decided for Z” but shall typically rationalize our choice after the fact.
Hi Dianelos,
ReplyDeleteI am not asking for unreasonable standards of rationality, not at all. I am simply pointing out the following:
First, the final call itself is the crucial part of your argumentation. Everything, I think, revolves around the possibility of the free agent making this call differently in different worlds.
Second, this final call, as you appear to recognize, is made entirely without conscious or rational supervision. Because if it were so, something could be said to explain it (even partially).
What happens before (and can be explained) or after (rationalization after the fact) may well be as you say. But, I think, this does not affect my point at all: moral responsibility appears to be based on a process (the final call) that is neither conscious nor rational.
jp
Dianelos
ReplyDeleteI'll give up on the coin toss analogy because for some reason you don't want to examine what I'm getting at by independence.
JP is absolutely right on this issue I think. The standard of rationality he requires is only the standard that yields moral responsibility.
By your example, in a deterministic world you do not steal a book because your ethics, needs, fears etc determine that you are not disposed to such actions. In your model, despite having an extremely strong aversion to the action, this does not determine your restraint. It is still entirely possible that you will steal, despite not wanting to (in the sense that there is only a 1 in 10 000) chance of you so acting because of your probability distribution). Indeed the requirement that your final calls appear random means occasionally such a thing must happen.
This is a very odd sort of freedom, one that insists that sometimes you do things you don't want to do at all, for no obvious reason at all, beyond 'I chose to and that's all there is to it', whereas most of the time you act in exactly the way you would were you living in a deterministic universe.
The problem of the uncaused cause is just that, it is uncaused, it floats free of the things I associate with moral responsibility (contemplation of the facts, implications etc).
Bernard
Hi JP,
ReplyDeleteYou write: “First, the final call itself is the crucial part of your argumentation. Everything, I think, revolves around the possibility of the free agent making this call differently in different worlds.”
Right.
“Second, this final call, as you appear to recognize, is made entirely without conscious or rational supervision.”
Let’s first consider this issue in the case of the random source model of decision making. That’s an entirely mechanistic model, so it’s easy to discuss. Now observe that what you write above is not true even in the case of the random source model. For even though the mechanical process which produces the final call uses a blind and dumb random source it is *not* the case that that mechanical process is made entirely without conscious and rational supervision. On the contrary, it is the interplay of the various rational and conscious factors (beliefs, desires, character etc) which determine that process, and therefore determine *how* the random source will be used.
But if your claim above is not true in the case of the mechanistic random source model, with more reason it is not true in the case of the personal agent causality model.
Similarly if the argument from the justification of different decisions based on identical past facts fails in the case of the random source model, with more reason it fails in the case of the personal agent model.
Nevertheless, let me revisit your whole point. You write: “Second, this final call, as you appear to recognize, is made entirely without conscious or rational supervision. Because if it were so, something could be said to explain it (even partially).”
Let’s again consider the random source model. What you evidently mean is that the random source event cannot be explained - which is true. But that random source event is just one part of the process which produces the final call. (The way I imagine it is that the various factors determine the probability distribution which instantiates a bag with colored marbles, one color for each possible choice. Then the random source is used to randomly, and hence blindly, pick one marble, in a process that runs subconsciously. The color of the marble determines the final call, say, if green then vote for the green candidate.) Now observe that one can rationally explain the instantiation of the bag; what one cannot rationally explain in the random source model is the random source event which determines the final call itself. Thus one can explain/justify most but not all of the decision making process. This last proposition holds both for the random source model and the agent causality model, but does not at all hold for the deterministic model, because there, as we saw, the whole explanation of why one has decided in a particular way is factually wrong.
Dianelos
ReplyDeleteIntriguingly I feel you have explained very well here the dilemma your model faces.
Yes, conditions determine the distribution of the marbles in the bag. So, to exactly this extent the process does indeed take into account those things we keep insisting upon; tastes, values, hormones etc. It can be expressed this way perhaps, to the extent the process is determinstic, it is responsive to the conditions.
But now we must choose a ball. If we do this blindly, we have a random model. If all the balls are the same colour, then the model is deterministic; probabilities have already collapsed and there is no freedom left.
By your model, you would say, I think, we choose a ball knowingly. Our free self now makes the call on which ball to pick.But observe, the function is probabilistic only to the extent that each ball has the same chance of being chosen, it is this random element that defines the probability function.
When I say your free will is random or blind, this is what I mean: every ball, now that their distribution has been determined, must have an equal chance of being picked. And this severely limits the nature of the free will you are proposing.
This is precisely why so many philosophers have concluded free will requires randomness, they are talking about the very independence you have identified in this example.
Beernard
Hi Bernard,
ReplyDeleteI agree with all you write in the beginning of your previous post. The process of decision making, which can take anything from seconds to months, is identical in both the random source and in the agent causality models, except for the very last but crucial moment where the actual final call is made. Here, according to the former model, an subconscious mechanisms kicks in which randomly (and thus blindly) picks a colored marble; in the latter model the agent (while keenly aware of the meaning of each color, and of why there are so many green and so few red marbles) consciously makes a sovereign decision and picks a color.
Now in the end of your post you write: “When I say your free will is random or blind, this is what I mean: every ball, now that their distribution has been determined, must have an equal chance of being picked. And this severely limits the nature of the free will you are proposing.”
Every ball (or “marble” in my-speak) does have an equal chance of being picked. But in the agent causality model this does not make it reasonable to call the agent’s free will “random or blind”. The agent, when it picks a ball, is not blind for it is keenly aware of the distribution of the balls and why it is as it is, and it picks not randomly for it does not use any random source but rather directly and consciously chooses. So, it is simply a fact about the model I am describing that the agent’s free will is neither random nor blind. Nevertheless you keep insisting that it is, or rather that it must be. On what grounds do you claim this? Why can’t it be the case that each ball has an equal chance of being picked, and also be the case that the agent freely picks one deliberately and consciously, and thus not randomly and not blindly? How do you get from “each ball has an equal chance of being picked” to “the agent’s free is random and blind”? It’s you who claim this, so it’s you who must justify your claim.
Dianelos
ReplyDeleteQuite. This is the crucial point. I am arguing that if each ball has the same chance of being chosen, then the knowledge about what the balls represent has not affected the choice. I am reasoning that if the information has affected the choice of balls, then by definition the chance of a given ball being chosen has been altered, so violating the requirement that each ball has an equal chance of being chosen.
This is the blindness I speak of. Any information about what the balls represent can not alter the probability that a particular ball is chosen. I think we both agree on this point. You don't like calling it blindness, I realise. Better perhaps to ask how moral responsibility can flow from this situation where the background information about the choice can not in any way sway the probability of one ball being chosen over another.
In other words, what does it mean for a choice to be directed, if it does not mean that the choice to which one is directed is in some way favoured? I can't make sense of this claim yet.
Bernard
Hi Dianelos,
ReplyDeleteI've been very busy the last few days and I only have a few minutes now, but here's something.
I am probably simply restating Bernard's point here but consider a number of possible worlds, identical down to the probability distribution. If the free agent is constrained by this distribution, it implies that over these worlds, the range of its different decisions will align itself on the probability distribution.
On way to do it is to replay the (possibly algorithmic) process that has produced the distribution but this seems excluded because the final call occurs after this process. Other than that and a random choice, how can the free agent act is such a way as to create, over all these worlds, the same probability distribution?
It seems to imply that, if not acting randomly itself, the free agent is severely constrained by a random process or some underlying influences (mechanism?) forcing it in various directions.
Or, to put it differently, is not being free and being forced to produce (over many worlds) results constrained by a probability distribution contradictory?
jp
Hi Bernard,
ReplyDeleteYou write: “I am arguing that if each ball has the same chance of being chosen, then the knowledge about what the balls represent has not affected the choice.”
Yes, but one should not try to understand agent causality through the lens of the random source model. At time T the two models work quite differently. In the random source model an actual set of colored balls (or something equivalent) must be instantiated for the mechanism of the random source to operate on. Not so in the agent causality model: Here the probability distribution *refers* to how the agent will choose at time T given the final balance of the various factors at play (beliefs, desires, character, external perceptions, etc). And the probability distribution has the form it has *because* the agent at time T is aware of these factors and their interplay, which thus influence its decision. The probability distribution is not something independent of the agent at time T, but rather is determined by the conscious state of the agent at time T (namely by the various factors at play). What the probability distribution does not say is what option the agent will choose, but it does say with what probability the agent will choose each option.
I hope I am making myself clear. It’s not like the agent must choose at time T in a way that does not violate the probability distribution. Rather the probability distribution simply describes how the agent, because of the various factors play, is apt to choose at time T.
“Better perhaps to ask how moral responsibility can flow from this situation where the background information about the choice can not in any way sway the probability of one ball being chosen over another.”
I am afraid it’s confusing to use language taken from the random source model when discussing the agent causality model. Re-reading my previous post I see that unfortunately I too have used such language. That was perhaps not wise, because it may keep you from correctly conceptualizing the agent causality model, and its key point that the probability distribution is not independent nor prior from the agent at time T, but rather describes a property of the agent’s state at time T. Incidentally, even if one sticks with the sub-optimal “colored balls” language when discussing the agent causality model, it’s not the case that the agent picks a ball, rather the agent picks a color.
Anyway, in the agent causality model the responsibility flows from the fact that no matter the probability distribution there is no power that may hold back the agent from picking what it will. For simplicity’s sake assume a situation where at time T the agent must choose between doing the wrong thing and the right thing, and given the various factors the respective probabilities are 75%-25%. If the agent chooses the evil thing then the agent has some responsibility for that choice because, even though the various factors made it less probable, the agent could have chosen the right thing if it had so willed, and it knew it was the right thing. (We are here ignoring the fact that the agent’s responsibility can be quite significant because the agent’s past choices have influenced many of the factors which in turn determine the probability distribution in the current situation.)
Dianelos
ReplyDeleteI understand your temptation to pull back from the ball analogy, because it does put in stark relief the contradiction at the heart of your model.
Either the probability distribution at time T describes the likelihood of each option being chosen at time T+1 or it doesn't. If it does, then the method by which the distribution collapses to a single choice (be this a mechanistic process or something far more mysterious) can not itself be informed by the circumstances that set up the distribution.
So, you still have to go back one step further and leave the notion of probability distributions out if you are really after a non-mechanistic explanation. Alternatively you could try to provide an alternative definition of probability distributions that allows an influenced and informed collapse that does not compromise the function's meaning. That won't be easy.
Or we can step away from your model entirely, now that it is no longer working for you, and note the same contradiction in far broader terms. An uncaused cause is by definition, well, uncaused. It is not moved, nudged, or informed by the circumstances, because at this point it is no longer sensible to call it uncaused. So, any part of the decision making process that is free in this sense will always be blind to the conditions. That's the problem. It's always been the problem, and clearly we could go around in circles for a very long time if you're not prepared to address the central dilemma.
Bernard
Hi Bernard,
ReplyDeleteYou write: “I understand your temptation to pull back from the ball analogy, because it does put in stark relief the contradiction at the heart of your model.”
As it happens our discussion has helped me understand agent causality much better than before, with not a trace of a “contradiction” in it. I pull back from the ball analogy because whereas the instantiation of a set of colored balls (or something equivalent) is necessary in the random source model, this is not the case in the agent causality model. Using such language when discussing the agent causality model may confuse one into thinking that the probability distribution is independent and prior from the agent’s free call, which is not here the case. When one thinks about the agent causality model one should not visualize the random source model; these after all are fundamentally different models.
“Either the probability distribution at time T describes the likelihood of each option being chosen at time T+1 or it doesn't.”
It does.
“If it does, then the method by which the distribution collapses to a single choice (be this a mechanistic process or something far more mysterious) can not itself be informed by the circumstances that set up the distribution.”
In the agent causality model there is no probability distribution prior to the agent’s final call to collapse at the time the final call is made. Neither is there a method the agent uses to make the final call. As you can see you keep using language (and therefore also concepts) from the random source model while referring to the agent causality model.
In agent causality it is not like the various factors determine a probability distribution which then the agent must respect when making the final call. Rather the agent, because of the various factors influencing it at the time it is about to make the final call, is apt to pick the various choices with different probabilities, and these probabilities shape the probability distribution we have been talking about. If anything then the probability distribution is a consequent of the agent’s final calling.
“An uncaused cause is by definition, well, uncaused.”
An agent is the uncaused cause of the final call in the sense that the free portion of the final call is not caused by anything beyond the agent’s own sovereign and irreducible will. And it’s because of this that the agent is entirely responsible for the free portion of its final call. Otherwise the agent could always argue that its beliefs, or its desires, or its character “made” it choose as it did, but such excuses now do not work, or at least do not fully work. That the agent has the power to be an uncaused cause of effects is what makes it possible for the agent to be personally and significantly responsible for its life’s path, and also for the agent to be a creator out of nothing.
I see now that understanding agent causality goes a long way in understanding theism, and indeed in understanding a basic sense of the theistic claim that we are made in the image of God.
Hi Bernard,
ReplyDeleteA few more comments on your last post.
You write: “It is not moved, nudged, or informed by the circumstances, because at this point it is no longer sensible to call it uncaused.”
Right. Nor would it be sensible to call it “sovereign” or call it “capable of creating out of nothing”.
“So, any part of the decision making process that is free in this sense will always be blind to the conditions.”
You keep suggesting a division (the conditions determining a probability distribution here, a method blind to these conditions determining the final call there) which is present in the random source model (for here the random source mechanism is indeed blind to the conditions and everything else for that matter), but which division is not in my description of agent causality. Moreover the division you suggest contradicts our experience of how it is like when the time comes and we make the final decision, for at that moment we experience being keenly aware of the various conditions.
Contradicting experience is a big no-no. Consider this: If naturalism is true then experience is contingent to physical facts and thus ontologically secondary. But even if naturalism is true experience remains epistemically primary, for all physical knowledge is inferred from our experience and is thus always provisional. In other words when the subjective/experiential narrative contradicts the objective/physical narrative it’s the latter that must be changed. Our experience of life is a sheer given brute fact, indeed it’s *the* given.
Dianelos
ReplyDeleteOur experience is a given, but not our interpretation of it. Our experience is entirely consistent with all three models, you have made this point yourself.
The probability function you describe under agent causality is no longer a probability function, in that the various likelihoods do not hold, even though in your last post you explicitly state they do. So the same contradiction keeps ringing clearly I'm afraid.
If you are able to explain what it means for probabilities to hold at time T under your agent causality model, then you may be able to begin building a coherent approach to free will, and in doing so revolutionise philosophy. It requires certain hubris I think to believe that you, and only you, can see this way forward, but good luck to you.
There is no hint of any such solution in your writing to date. Whether you are wilfully avoiding this for the sake of your prior beliefs only you can say.
To repeat, what on earth do you mean by probabilities existing at time T?
Bernard
Hi Bernard,
ReplyDeleteYou write: “ Our experience is entirely consistent with all three models, you have made this point yourself. ”
True, in the sense that in the random source model one can visualize a state of affairs where our brain is massively fooling us into believing that it is we who (while brightly aware of the various factors that affect us) are making the final call, whereas in fact it is a completely blind mechanism that does that. But in the context of agent causality to insist, as you do, that when we agents make the final call we are blind to the factors (beliefs, desires, character traits) that influence is, flatly contradicts our experience. What I was saying then is that it’s your particular description of agent causality which contradicts experience, and thus cannot be true.
“To repeat, what on earth do you mean by probabilities existing at time T?”
To use an example you introduced before: A voter ponders whom to vote among two candidates, the green one and the union one. The various factors at play (personal ones such as beliefs, desires, character traits, as well as external factors such as what he reads in the papers) evolve during various days thus shifting the probability that in the voting booth the voter will pick one or the other candidate. Finally, at time T, the voter is in the voting booth and has to make his final decision. At that time the final state of the various factors that pull and push on the voter's conscience is such that the probability that the voter will vote green is 75% and that he will vote red is 25%. (Incidentally, these probabilities have a nice mathematical meaning: The best strategy for betting about how the voter will vote is by assuming precisely these probabilities.) Finally, at time T, and even though the various factors (of which the voter is always brightly aware) are overall pulling the voter towards the green candidate, the voter decides to vote union.
The above account strikes me as so ordinary (both in its experiential and in its conceptual content) that I would appreciate you explained what you find contradictory or incoherent in it.
Hi Bernard,
ReplyDeleteOne more comment. You write: “It requires certain hubris I think to believe that you, and only you, can see this way forward, but good luck to you.”
I’d say it requires a certain independence of thought, rather than hubris. Anyway I know for a fact that many philosophers (the so-called libertarians about free will) do not see any contradiction or incoherence in the concept of free will, so I am certainly not alone in this. I don’t know though if their view of how free will fits reality is the same as mine. I have started reading the book about free will I have told you about, and I expect to soon find out.
Dianelos
ReplyDeleteThere is no contradiction in your example if you assume the final decision to vote union is made without reference to the conditions that created the 25/75 split. This is the random model and it is entirely coherent, albeit without any solid scientific support at this point.
If however you insist that the agent's awareness of the relevant factors influences the decision to vote union, then the contradiction is clear, the 25/75 can no longer hold because the factors now driving the decision are properly part of the probability function, and the decision collapses to the deterministic model, where the factors themselves create the probability at T+1, the 100% probability that a union vote will occur.
Now, you tend to escape the contradiction in the second case by claiming that no, in fact the agent is not swayed by the conditions, but has sovereign power. When it is pointed out that this ignoring or over riding of the conditions amounts to randomness, you defend this by retreating to the second case, arguing that no, the agent is aware of all the factors and takes these into account when making the final call.
This is in essence a debating trick, running two separate lines and defending the weakness in each by claiming that this isn't really your model. This is why you didn't like the balls analogy, despite it being your own contribution, because it forced you to choose one of the models, and none of them worked for you.
The question then is this, is the agent influenced by the conditions underpinning the probability function
when collapsing in favour of union? This is a simple yes/no question. Answer this categorically and we can attempt to build your model from there. I would love you to succeed in this by the way, it would be a most exciting result.
Bernard
Hi Bernard,
ReplyDeleteI missed this question of yours: “Is the agent influenced by the conditions underpinning the probability function collapsing in favour of union?”
Yes. But it bothers me that you continue to use language which is foreign to the model I am describing, and which language I find somewhat opaque. Why not simply ask: “Is the agent influenced by the conditions that make it more probable that it will vote green, at the time when it decides to vote union?” Actually I hope that that’s the question you meant to ask, because that’s the question I am answering.
Dianelos
ReplyDeleteI'm sorry that my choice of language causes you difficulty here. I'm just trying to follow through the implications of the probability distribution model, which is yours not mine. I'm trying extremely hard not to be opaque, and will continue to do so.
Okay, so you say the agent is influenced by the conditions. Note this is different from saying the conditions define the probability function. That's a given and their role in doing so is deterministic. So we're interested in what, if anything, is left for the free self to do once deterministic forces have set up the probability function. You seem to be saying the free self makes a sovereign call with reference to the conditions.
Now, I think your claim that the agent is influenced by conditions creates a contradiction. If there is a 75/25 split and this is a meaningful measure, then we're saying that faced with identical conditions, sometimes (25% of the time) the agent will go union, the rest of the time green. Under exactly the same conditions, two different outcomes can occur.
In what sense then (apart from obviously setting up the 25/75 ratio) is the agent influenced by these conditions, if under precisely the same conditions he/she can go either way? That doesn't seem like any influence at all. I can make no sense of this.
Bernard
[continued from above]
ReplyDelete“In what sense then (apart from obviously setting up the 25/75 ratio) is the agent influenced by these conditions, if under precisely the same conditions he/she can go either way?”
In no other sense. The only way the conditions influence the agent’s will (or to be more precise the agent’s final deciding) is by making it more probable that the agent at time T will pick one choice rather than the other (in exactly the 25/75 ratio). That’s as far as the conditions go. From here on the agent’s free, sovereign, and creative will kicks in to determine the final call. The language above though must be understood logically and not temporally. It would be a misleading to think that the agent’s free will must respect the 25/75 ratio because this distribution is set before it makes the final call. Rather the agent’s will is such that, given the conditions and their influence, it will determine the vote for union or green at 25/75 probabilities. Or, if you prefer to use the physical narrative: At time T the relevant state of the agent’s brain is such that, given the conditions and their influence, it will determine the vote for union or green at 25/75 probabilities.
In this context please observe that the concept of probability does not refer to states but to events. When, for example, we say that when spinning a fair coin we get heads with probability 50%, that probability is a property of the event of spinning a fair coin, and not of the final state of the coin. We say there is a 50% of getting a head *given the event of spinning a fair coin*. In contrast, spinning an unfair coin is a different event, and here getting a head may have a probability of 60%. What changed is not the final state of the coin (it’s always a head) but the event. When speaking of a probability distribution in the context of either the agent causality model or the random source model we mean a property of the event that produces the final call, in the former case the mechanism of a random source applied to a set of colored balls, in the latter case the non-mechanical free will of the agent. And that probability describes to what degree the agent’s deciding is influenced by the various conditions, but does not describe how the agent will finally decide.
Bernard, even though our discussion has been quite fruitful for me so far, I should be investing more time in other tasks. So I’d like from now on to post mainly on weekends. Meanwhile I shall continue reading the book about free will (I read philosophy every night because it puts me to sleep). I have already discovered some interesting facts about the modern philosophical discussion on free will, which I shall describe here in good time.
Dianelos
ReplyDeleteExcellent. So the agent is influenced in 'no other sense', the only influence the conditions have is in the setting up of the probability distribution. Exactly as it is for the random model. So, the final call is not only uncaused, it has no rational component.
An example. I, despite the 25/75 split, use my free will to vote union, and meet a friend as I leave the voting booth. Consider the following conversation:
-Who did you vote for?
-Union.
-I thought you were leaning green.
-I was.
-What changed your mind?
-I just did.
-Did you prefer their policies?
-No, I preferred Green policies.
-Did you like the union candidate more?
-He's a good man, but no, I liked the Green better.
-Were you voting tactically?
-No, the green vote would have sent a clearer message.
-Were you just overcome by an urge to vote union?
-No, at the moment of voting, I could feel a biochemical urge to vote green.
-So why did you vote union?
-No reason. I was exercising my free will.
Once you strip the final call of all reference to the conditions (as you must if in an identical world your free will can act otherwise) you end up with an empty vessel. There's no moral responsibility to be found here. The model is simply vacuous.
This is not a difficult point. Only you can know why it is you have resisted it so vigorously.
Bernard
Dianelos
ReplyDeleteI've just been reading a chapter on the problem of free will from Thomas Nagel's 'The View from Nowhere.' I think he describes the problem pretty well, if you still have any appetite for the topic after your current reading.
Bernard
Hi Bernard & Dianelos,
ReplyDeleteTo summarize: as far as I understand what Dianelos is saying, I see at least three (somewhat interrelated) issues with agent causality. (Bernard: tell me if I am missing something important here.)
First, the probability distribution. The existence of this (entirely determined) distribution implies that, over a large number of possible worlds, the set of final calls must comply with it. A random selection fits perfectly the bill here but it is not clear at all how the free agent can be constrained in such a way as to reproduce the distribution.
Second, the rationality of the agent. As every argument based on the facts of the case must be expressed using them and because these are the same in all worlds, the agent can offer no rational explanation at all for any of its free choice (as illustrated by Bernard in his last comment).
Third, moral responsibility. Because of the above, delegating moral responsibility to the free agent seems to be a most “irresponsible” thing to do (if I may).
The answer to all that (and to various questions of why and how) is to posit the existence of a free agent, irreducible, sovereign, acting without cause just in the manner necessary to produce the needed result. In other words, free will is explained by asserting the existence of a part of us that is, well, free. Seems like circular reasoning to me.
Dianelos: we are probably not looking for the same kind of answer here. Maybe you only goal is to build a logically consistent framework that leaves a place for free agency. If so, you may possibly succeed, I don't know – I suppose anything can be made to fit such a framework if we make the right assumptions. The problem, as I see it, it not inconsistency but how many of these loose ends, unanswered questions and ad hoc assumptions we can accept. The issues I summarize above are real and saying “it is simply so” is hardly a satisfying answer.
jp
JP
ReplyDeleteI think the whole thing does come down to intuitions which themselves are beyond analysis. I've been reading around the topic a bit and the philosophers I most chime with take such a point.
So to you and me, it just seems obvious that Dianelos' version of free will doesn't yield a satisfactory type of moral responsibility. And yet to Dianelos, and many others, it seems equally obvious that the compatabilist claim that determinism can yield a satisfactory view of freedom just doesn't work at all.
And actually the more I dwell on it the more I am inclined to agree with him here. I don't think the materialist view provides any sort of freedom that I can find meaningful. This is an entirely personal intuition, just as my sense that Dianelos' model is deficient is a personal intuition.
And so, predictably, I find myself in the middle, suspicious of both solutions because they appear to me to be to be dancing around the very heart of the matter; which is how I feel about both theism and atheism in general.
And so I must read and think further.
Bernard
Bernard,
ReplyDeleteI am not sure my issue with Dianelos' approach is primarily a question of intuition. Rather, I think, it is because I am looking for different kinds of answer, so that what is satisfactory for him might not be so for me and, most likely, inversely. I am looking forward to his take on this and it would be most interesting to clarify those different requirements, so to speak.
I am somewhat confused about the role of intuition in general. Many philosophers, so it seems, rely a lot on intuition. In this respect, philosophy seems to be at a crossroad between science, where intuition as a source of knowledge is a no-go, and art, where intuition is essential (of course, scientists and mathematicians use intuition all the time as a guide but it is seen as eminently unreliable). I wonder how this relates to the connection you have often mentioned between religion and art.
As for free will, I think one problem (at least, it seems so to me) is that we don't really know what we are talking about. We've been discussing this for weeks on this blog and the “what” we are looking for seems as elusive as ever. I can understand freedom to some degree as a mechanical or algorithmic process but I must admit that the more I try to understand what Free Will is supposed to be, the less it makes sense. Maybe this search if for a ghost, a shadow without substance. Maybe a large part of the question is simply semantic and that, once we have cleared up the ambiguities and imprecisions, we end up with a notion of free will that is both intelligible and satisfactory. For what it's worth, this is my “intuition”.
Hi JP
ReplyDeleteMy comments here keep dropping off. I did try to respond a couple of times. This business of the role of intuition is important and I find it difficult. I may find time to try again.
Bernard
Hi Bernard,
ReplyDeleteI got your comments by email. Those who didn't comment on this post on didn't subscribe to email followups will not have seen them however. I suppose they are lost in spam territory, as Eric explained a while ago. Maybe he could check and see if he could restore them.
Yes, these are difficult questions and quite confusing. I will probably have time to comment over the week-end.
JP
Bernard,
ReplyDeleteYou write: “An example. I, despite the 25/75 split, use my free will to vote union, and meet a friend as I leave the voting booth. Consider the following conversation:
-Who did you vote for?
-Union.
-I thought you were leaning green.
-I was.
-What changed your mind?
-I just did.
-Did you prefer their policies?
-No, I preferred Green policies.
-Did you like the union candidate more?
-He's a good man, but no, I liked the Green better.
-Were you voting tactically?
-No, the green vote would have sent a clearer message.
-Were you just overcome by an urge to vote union?
-No, at the moment of voting, I could feel a biochemical urge to vote green.
-So why did you vote union?
-No reason. I was exercising my free will.”
Two observations: First people usually rationalize their choices, so that’s not how people speak. I assume this is a dialogue with a philosopher who is a libertarian about free will and who wants to describe his experience at time T as coherently as he can. Secondly, the dialogue you suggest where everything is in favor of voting green does not concord with a 25/75 split but rather with a 1/99 split. In a realistic 25/75 split there would be some important factors in favor of voting union. Often there is a conflict between reason and emotion, so I’d suggest that the following dialogue is more realistic:
-Who did you vote for?
-Union.
-I thought you were leaning green.
-I was.
-What changed your mind?
-Nothing changed my mind; rather I made up my mind.
-But did you prefer the union policies?
-No, I prefer green policies.
-Did you like the union candidate more?
-Actually, yes, I know her personally and she is a good person. Normally though I let logic rather than emotion determine my choices, but not this time.
-Why not this time?
-No particular reason. When the moment came to make a decision, I decided to endorse the emotional reasons at hand and so I voted for the union candidate.
[in the next post I suggest how a determinist about free will and how an indeterminist about free will would respond]
[continued from above]
ReplyDeleteNow let’s suppose that the philosopher voter is a determinist about free will. The dialogue that would ensue would be something like this:
-Who did you vote for?
-Union.
-I thought you were leaning green.
-I was.
-What changed your mind?
-Nothing changed my mind; my mind or rather my brain was going to vote union all the time.
-But didn’t you say you were leaning green?
-I meant I thought I was leaning green. We are often not aware of what’s really happening in our brain.
-But didn’t you prefer green policies?
-That’s difficult to say. It certainly seemed to me I preferred green policies, and perhaps I really did, on the other hand other emotional factors which seemed to me to be secondary turned out to be stronger. Or perhaps there are factors I am not aware of which pushed the balance towards union.
-Sounds like you are not in control of your own decisions.
-On the contrary my brain is firmly in control of its decisions, only my awareness of my brain is feeble and often confused. If it weren’t I would know beforehand that I would vote union, but in fact entering the voting booth I still thought there was a 75% probability that I would vote green.
And here is how a philosopher who believes in indeterminism about free will would describe his decision:
-Who did you vote for?
-Union.
-I thought you were leaning green.
-I was.
-What changed your mind?
-Nothing changed my mind; my mind or rather my brain is an indeterministic machine, so it was really randomness which made up my mind.
-But didn’t you say you were leaning green?
-I was, and that’s why the wavefunction of my brain would with a 75% probability collapse into a physical state that would cause me to vote green. But, as it happens, the wavefunction collapsed the other way.
-But didn’t you prefer green policies?
-I did, and this preference as well as the interplay of all other relevant factors – for example I personally know the union candidate and she’s a good person – build up the exact shape of my brain’s wavefunction. It was more probable that I would vote green, but, as chance would have it, I voted union.
-Sounds like you are not in control of your own decisions.
-I am up to the shape of the wavefunction. The rest is chance.
Hi Dianelos
ReplyDeleteMy comments no longer seem to appear on this thread so perhaps this is a good reason to leave this for now.
I have been reading an article by Robert Kane who appears to be taking a similar line to you. Very interesting although I don't yet fully understand it.
Bernard
Bernard,
ReplyDeleteWhat I find helps is not to write long posts (less than 2,000 characters if possible) and also not to cut and paste in the "Leave your comment" window and then immediately click on the "Publish your comment" button. I assume that google has implemented heuristics to filter spam which dislike long posts as well as posts that are published too quickly. So, even though I prepare my posts in a word processor, after I copy them into the "leave your comment" window I wait a few minutes before publishing them.
Anyway I do get your posts by email. If you wish I could try and publish them into Eric's blog myself.
Hi Bernard,
ReplyDeleteI am not sure origination is incompatible with determinism. You are right to say, I think, that answering it is incoherent and that nothing is lost if we can't have it is not satisfactory. But there may be other ways to look at this.
On aspect of the question I think is important is that these two notions, determinism and origination (or free will), are applicable to completely different things. It is not at all obvious to me how we can put them together in some meaningful opposition. At the same time, while determinism seems rather well defined, this is not the case at all with free will. Another problem may be that the word “determinism” has connotations of “pre-decided” or “forced”. This is clearly not the case – no outcome is explicitly decided in advance. I fear that these connotations often trump the underlying meaning and serve only to confuse the issue (at least for me). The term “algorithmic” is probably more useful.
[...]
[...]
ReplyDeleteThe central issue, then, is whether an algorithmic process can produce origination. Algorithmic means only that the process follows rules (any rules) and it may be deterministic or not, depending on the presence of random elements. Now, to establish that origination is incompatible with an algorithmic process we need to do two things: first, define precisely an example of an outcome that would qualify and, second, show that no algorithm can produce it. This is often assumed but I think it is far from easy to do. Simply defining what we're trying to is not obvious at all (at least to me).
The predictability issue seems a red herring to me. We have discussed this before and it may not be feasible to actually predict the future (with 100% accuracy) by simulating less than the whole universe. But, in any case, predictable may mean no more than this: let events unfold and something will happen. I have a hunch that this thing about different outcomes in different worlds that free will proponents use all the time might be an overkill. Well defined (can it be done?), this condition may be sufficient to establish free will but it does not mean it is necessary. Likewise with non-predictability.
jp
Hi Dianelos,
ReplyDeleteI find it striking that today, when is as scientifically certain as it gets that determinism is false, many people (e.g. Dennett) still insist with a deterministic understanding of free will.
I think this is done simply to simplify the argument. If free will and personal responsibility can be crafted out of determinism then, as randomness cannot destroy it, this is also true of reality. Complete determinism represents the “worst case”, if you wish. In this context, random events don't add anything useful. They are just so much background noise and it's simpler to simply ignore them.
JP
ReplyDeleteYou're right, defining origination appears central, because at this point we can meaningfully ask whether it is compatible with determinism. Origination itself stems from an intuition, a personal need in terms of one's own story, perhaps. If I am to hold hopes, be proud of or disappointed in my own efforts, then there must be a meaningful sense in which they are my own. This appears to require some type of origination. But put this way it is cloudy and ghostlike. My guess is the definition is hard because what each of us needs to complete the story of personal responsibility may be different.
My personal instinct is that origination means that at time T, before a decision is made, the complete information required for the algorithm to play out is not in existence. Some aspect, that originates with me, is added in order for the decision to be made. Determinism, as I understand it, says all the information is in place at time T, even if we can't use it to generate an accurate prediction. My hunch is that decision making is deterministic, I think of some of the compromised will effects of split brain patients for example, but under this definition, origination does seem to be lost.
Are you saying there might be a way around this apparent conflict, or are you saying that we need to craft new metaphors to split personal agency off from this kind of origination?
Bernard
(reposted)
ReplyDeleteHi Bernard,
You write: “Although we both might tweak the conversations for the purposes of spin (I'd not have determinists speaking about 'my brain' as separate from 'me' for example) under any version the key problems remain for me inescapable.”
Actually I found that the exercise of devising such dialogues is quite useful. I did tweak the libertarian dialogue, for I wanted to make it as much justice as possible. Feel free to similarly tweak the other dialogues. I wonder, how would Daniel Dennett do it?
You say that a determinist does not speak of “my brain” separate from “me”, but I understand that is a linguistic shortcut and is not meant literally. After all “my brain” does a lot of things of which “I” am not at all aware and I am certainly not doing, so the two can’t be identical. What I understand the determinist is saying that “I” is identical to some *part* of “my brain”. But then, and that was a point that came out in the dialogue, the part of my brain which is identical to me is not being well informed by the rest of my brain. Or, not to mince words, the rest of my brain is continuously and massively fooling the part of my brain which is me. After all, when the deterministic philosopher entered the voting booth he still believed that there was a 75% probability that he would vote green when in fact that probability was zero.
“The libertarian is left with at some point 'I just decided to, no reason' and the determinist must include something like 'I did what I was always going to do.”
To me “I just decided to, no reason” may sound a little awkward, but in fact states the obvious, namely that “I” decided and not the “reasons”. Any account where in the end or to some degree the decision is *mine* must be such that there is no “reason” that caused it. I have finished the essay in the book “Four Views on Free Will” and will copy here some relevant points from it.
[continued next]
[continued from above]
ReplyDelete“Now both groups intuitively feel their particular compromise is of no great moment, whilst the other's is fatal. A great amount of ink can be spent attempting to show the weakness doesn't exist at all, but eventually we must retreat to showing why the weakness is not important. And this does seem to be more about intuition than reason, I can't see how either group makes the next move of showing why their vulnerability doesn't matter, beyond appealing to intuition.”
I don’t think it’s a question of intuitions, as the dialogues above demonstrate. While the libertarian dialogue is kind of awkward, the random source dialogue and especially the deterministic dialogue strike me as bizarre, and I don’t think they are fixable.
I find it striking that today, when is as scientifically certain as it gets that determinism is false, many people (e.g. Dennett) still insist with a deterministic understanding of free will while loudly advertising their scientific pretensions. The second essay in the abovementioned book is about compatibilism and it’s interesting how the issue is defined: The author doesn’t actually claim that determinism is true, but claims that if determinism were true then we could still make sense of free will and personal responsibility. It’s like somebody assuming a contortionist position just to prove that it can be done. I don’t get it.
“I have been reading an article by Robert Kane who appears to be taking a similar line to you.”
Robert Kane wrote the libertarian essay in the abovementioned book. For my taste it is more complicated than necessary, and his intent to make sense of free will without introducing some transcendental non-mechanistic ingredient (which appears to be the norm among libertarian philosophers) strikes me as unnecessary and moreover as something that can’t possibly work.
Hi JP,
ReplyDeleteYou write: “If free will and personal responsibility can be crafted out of determinism then, as randomness cannot destroy it, this is also true of reality.”
Free will and personal responsibility is not something to be “crafted” but something that must “fit” within one’s understanding of reality. And as we now know that physical reality is not deterministic, I really don’t see what profit there is for a physicalist to try to show that free will and personal responsibility are compatible with something that does not actually exist.
“Complete determinism represents the “worst case”, if you wish. In this context, random events don't add anything useful. They are just so much background noise and it's simpler to simply ignore them.”
I disagree. Let’s look at this matter from a scientist’s point of view. For a scientist our brain is a machine, and free will is observed by its consequences namely the relevant part of that machine´s behavior. Now, conceptually, there are deterministic and indeterministic machines, but in physical reality there are only indeterministic machines. Indeed, we have to invest a lot of effort to build actual machines that *approximate* a deterministic machine, such as clocks or computers. (The reason we don’t build smaller or faster computer chips is that they begin to behave indeterministically.) And today we know enough about our brain to know that the probability that our brain happens by nature to approximate a deterministic machine is virtually zero. So indeterminism (or “random events” as you put it) are *not* background noise but a fundamental property of our brain’s behavior.
Hi Bernard,
ReplyDeleteI find (in my case at least) that much of the difficulty of this issue is psychological. I clearly feel the pull of the idea of determinism towards this view that everything is decided, that all notion of free will is just an illusion. But, when I try to look deeper into this, I can't really put the two together. Something is just wrong – these two don't belong in the same place. A mélange des genres of sort.
On the other hand, when I think algorithmic process – meaning a process following rules – it does not look so bad. This idea (more accurate in any case) does not provoke these associations that determinism does, while allowing for the non-determinism of reality that Dianelos finds so important. The question then would be: what is it, specifically, that an algorithmic process cannot accomplish?
I think a crucial point is that free will remains a very fuzzy notion. It is always approached obliquely by specifying a condition that is assumed equivalent. For instance, consider the idea of (potentially) getting different results in different worlds. This if often seen as equivalent to free will but, while it may perhaps be a sufficient condition, it does not follow that it is necessary. Likewise, what you require for origination may very well be sufficient but not necessary.
We say free will requires this or that but I suspect this is in large part out of habit. I find it quite difficult to pinpoint what is really necessary. But what we need, and even what we really want, may not be what it seems to be. The key may be to find out a way to untangle all this and, hopefully, see more clearly.
By the way, I don't find Dennett very persuasive on this and I have told you before of my initial reaction to his take on free will. But what stayed with me is this: the guy was not trying to build a makeshift free will out of the determinist wasteland (which at the time seemed to me the only viable option). No, he was going squarely after the big game, as if the contradiction between an algorithmic process and free will was no more than a house of cards and that, by letting some fresh air in, it would collapse into nothingness. In this, I think he may be ultimately right.
Hi Dianelos,
ReplyDeleteYou intrigue me. Are you saying that an argument for free will in a deterministic reality would not apply to a combination of determinism and randomness? In what way would randomness undermine the argument?
Dianelos
ReplyDeleteWhen you write:
"...the random source dialogue and especially the deterministic dialogue strike me as bizarre,"
you are surely making an intuitive argument. Others find the "I just did it, no reason" bizarre, and that's why the debate goes on.
JP may be right, that the trick is untangling further what we want from free will and why. Most writers proceed as this point is just obvious and maybe that is the problem.
Bernard
JP
ReplyDeleteThis is a good question. What are the necessary conditions for free will? The answer will depend upon what one uses the concept for, so for instance some theists might require a particular version of free will in order for it to be consistent with their doctrine of personal responsibility.
The question then becomes, for anyone interested in the free will problem, what does each personally require? I'll need to think about this. I know it has something to do with concepts such as regret, loyalty, generosity or satisfaction, which seem to me to be woven into the very fabric of our stories of self.
So what are the necessary conditions for such concepts to be coherent? I don't honestly know, it feels like a very difficult question.
For now I suspect origination is going to be the stumbling block. One way to approach it is to try to describe an algorithmic model where something like generosity still makes sense. In this regard, I don't think Dennett is quite there yet. As with consciousness, he is floating an hypothesis: 'maybe, when we examine this thing closely, the mystery will disappear'. Maybe.
Bernard
Hi JP,
ReplyDeleteYou write: “Are you saying that an argument for free will in a deterministic reality would not apply to a combination of determinism and randomness?”
In the physical reality, as described by science, there is no such thing as a “combination of determinism and randomness”. I mean it’s not like on takes two properties which do not describe reality, mixes them, and gets something which does describe reality. In any case the whole issue is an unnecessary complication. According to what we now know beyond reasonable doubt all physical systems evolve indeterministically. And “indeterminism” means “not determinism”.
Determinism means that given the current physical state or a system there is only one possible future state that it may evolve to. Indeteterminism means that given the current physical state of a system there are more than one possible future states that it may evolve to. Actually science tells us that there is a continuum of a practically infinite number of future states, described by the so-called probability distribution. Nevertheless in large and simple physical systems (such as dropping an apple, or a working car engine) the probability distribution is so concentrated around one future state (say, the apple will fall downwards) that it is practical to analyze it as approximately deterministic. Still, modern science tells us there is a minuscule probability that the system will evolve towards a completely different future state (say, the apple will fly upwards as if by magic).
Given what we know about the brain (both because of its complexity and because of the quantum mechanical nature of critical events such as the movement of ions along synapses) the probability that a human brain can be studied as an approximately deterministic system is virtually zero. (This according to my understanding; do you know of any serious scientific source which claims that the brain will turn out to behave as an approximately deterministic system?) Which means that, happily enough, science confirms a basic sense we have about our actions, namely that in many cases where we must choose among two alternatives, both can be chosen with some significant probability. Or, to put it in other words, the relevant probability distribution is such that both future states display a significant probability peak.
Hi Dianelos
ReplyDeleteMy favourite take on quantum weirdness is that version of the Copenhagen interpretation that says, in essence, 'there is only the measurement. Of the unmeasured state, there is nothing that can be said.'
Your take is a little different to this, and it's good to be aware I think, when dealing with Quantum stuff, that the interpretations remain deeply controversial and unresolved. Even a statement such as, 'we now now the world is not deterministic' relies upon a specific and as yet unestablished interpretation.
The relevant question for brain science then will not be 'is the process deterministic' so much as, 'if quantum effects are important to decision making processes, in what way are they important?' And here, as far as I know, we are a huge way away from a sensible, detailed answer.
Bernard